chapter 9

WHAT DOES IT TAKE TO GET THESE BEHAVIORS? WHAT THEORY SAYS

Another way of phrasing these same questions is to ask, "What motivates employees?" For example, we know that people differ in the importance they attribute to money. That is partly because people differ in the value they place on money and partly because the same person may place a different value on money depending on their circumstances (e.g., their age, wealth, aspirations).13 We all know someone who just isn't motivated by the almighty dollar. But even for these people, it's possible to design a pay system that is appreciated—one in which the link between effort and pay is evident. Although money is not everyone's prime motivator, money is, on average, the most important reward in surveys of employees. Also, whatever money's importance, given most people's strong preference to be treated fairly, almost nobody likes to be underpaid or reacts well to it. Is it any wonder, then, that figuring out the motivation equation is a major pastime for compensation experts?14 In the simplest sense, motivation involves three elements: (1) what's important to a person, and (2) offering it in exchange for some (3) desired behavior. As to the first element, what's important to employees, data suggest most employees prefer pay systems that are influenced primarily by individual performance and the market rate, with seniority also important to some.15 To narrow down specific employee preferences, though, there has been some work on what's called flexible compensation. Flexible compensation is based on the idea that only the individual employee knows what package of rewards would best suit personal needs. Employees who hate risk could opt for more base pay and less incentive pay. Trade-offs between pay and benefits could also be selected. The key ingredient in this new concept is careful cost analysis to make sure the dollar cost of the package an employee selects meets employer budgetary limits.16 Absent widespread adoption of flexible compensation systems, we need to answer these three questions the old-fashioned way—by going back to theories of motivation to see what "makes people tick." In Exhibit 9.4 we briefly summarize some of the important motivation theories.17 These theories try to answer the three questions we posed above: What's important? How do we offer it? How does it help deliver desired behaviors? Pay particular attention to the "So What?" column, in which we talk about the theories' views on how employee behavior is delivered. Some of the theories in Exhibit 9.4 focus on content—identifying what is important to people. Maslow's and Herzberg's theories, for example, both fall in this category. People have certain needs—such as physiological, security, and self-esteem needs—that influence behavior. Although neither theory is clear on how these needs are offered and how they help deliver behavior, presumably if we offer rewards that satisfy one or more needs, employees will behave in desired ways. These theories often drive compensation decisions about the breadth and depth of compensation offerings. Flexible compensation, with employees choosing from a menu of pay and benefit choices, clearly is driven by the issue of needs. Who best knows what satisfies an employee's needs? The employee! So let employees choose, within limits, what they want in their reward package. Theories of a second sort, best exemplified by expectancy theory, equity theory, and agency theory, focus less on need states and what rewards best satisfy those needs and focus more on motivational processes, including how perceptions of needs and other factors (equity/fairness, risk, linkages between effort, performance, and pay) are processed cognitively to determine motivation and behavior.18 Many of our compensation practices recognize the importance of a fair exchange. We evaluate jobs using a common set of compensable factors (Chapter 5) in part to let employees know that an explicit set of rules governs the evaluation process. We collect salary survey data (Chapter 8) because we want the exchange to be fair compared to external standards. We design incentive systems (Chapter 10) to align employee behavior with the needs (desired behaviors) of the organization. All of these pay decisions, and more, owe much to understanding how the employment exchange affects employee motivation. Expectancy theory argues that people behave as if they cognitively evaluate what behaviors are possible (e.g., the probability that they can complete the task) in relation to the value of rewards offered in exchange. According to this theory, we choose behaviors that yield the most satisfactory exchange. Equity theory also focuses on what goes on inside an employee's head. Not surprisingly, equity theory argues that people are highly concerned about equity, or fairness of the exchange process. Employees look at the exchange as a ratio between what is expected and what is received. Some theorists say we judge transactions as fair when others around us don't have a more (or less) favorable balance between the give and get of an exchange.19 Even greater focus on the exchange process occurs in the last of this second set of theories, agency theory.20 Here, employees are depicted as agents who enter an exchange with principals—the owners or their designated managers. It is assumed that both sides to the exchange seek the most favorable exchange possible and will act opportunistically if given a chance (e.g., try to "get by" with doing as little as possible to satisfy the contract). Compensation is a major element in this theory, because it is used to keep employees in line: Employers identify important behaviors and important outcomes and pay specifically for achieving desired levels of each. Such incentive systems penalize employees who try to shirk their duties by giving proportionately lower rewards. At least one of the theories summarized in Exhibit 9.4 focuses on the third element of motivation: desired behavior. Identifying desired behaviors—and goals expected to flow from these behaviors—is the emphasis of a large body of goal-setting research. Most of this research says that how we set goals (the process of goal setting, the level and difficulty of goals, etc.) can influence the performance levels of employees.21 For example, workers assigned "hard" goals consistently do better than workers told to "do your best."22 A final theory we will mention (not shown in Exhibit 9.4) purports to integrate motivation theories under a broad umbrella. Called self-determination theory (SDT), this approach believes that employees are motivated not only by monetary rewards (referred to as extrinsic motivation), but also by intrinsic motivation, which is enjoyment or satisfaction that comes from performing the work itself and produces a sense of autonomy. According to SDT, intrinsic motivation produces the highest-quality motivation.23

DOES COMPENSATION MOTIVATE BEHAVIOR?

Now let's look at the role of compensation in motivating the four types of behavior outlined earlier: the decision to join, to stay, to develop skills, and to perform well. Do People Join a Firm Because of Pay? - Level of pay and pay system characteristics influence a job candidate's decision to join a firm, but this shouldn't be too surprising.41 Pay is one of the more visible rewards in the whole recruitment process. We know, for example, that high-ability applicants do select companies because they provide pay for good performance.42 This so-called sorting effect has been shown both for good employees choosing a company and for bad employees leaving a company—all because they do (or do not) like the pay system being used.43 Job offers spell out the level of compensation and may even include discussions about the kind of pay, such as bonuses and profit-sharing participation. Less common are statements such as "You'll get plenty of work variety," or "Don't worry about empowerment," or "The workload isn't too heavy."44 These other rewards are subjective and tend to require actual time on the job before we can decide if they are positive or negative features of the job. Not so for pay. Being perceived as more objective, it's more easily communicated in the employment offer. Recent research suggests job candidates look for organizations with reward systems that fit their personalities.45 Below we outline some of the ways that "fit" is important. None of these relationships is particularly surprising. People are attracted to organizations that fit their personalities. Evidence suggests that talented employees are attracted to companies that have strong links between pay and performance.51 One way to get this linkage is to give employees some control over the rewards they received. Thirty years ago no company ceded this control to employees. Now almost all major companies allow employees to have some reward choice. And the impact of reward choice is positive: up to 40 percent improvement in performance but only if the choices available are attractive to employees.52 It's not a big jump, then, to suggest that organizations should design their reward systems to attract people with the desired personalities and values. For example, if we need risk takers, maybe we should design reward systems that have elements of risk built into them. Do People Stay in a Firm (or Leave) Because of Pay? - Employee decisions to leave are influenced by their performance and the degree to which pay is performance-based.53 How does pay affect this relationship? Much of the equity theory research in the 1970s documented that workers who feel unfairly treated in pay react by leaving the firm for greener pastures.54 This is particularly true under incentive conditions. Turnover is much higher for poor performers when pay is based on individual performance (a good outcome!). Conversely, group incentive plans may lead to more turnover of better performers—clearly an undesirable sorting effect.55 When AT&T shifted from individual to team-based incentives a number of years ago, star performers either reduced their output or quit. Out of 208 above-average performers, only one continued to report performance increases under the group incentive plan. The rest felt cheated because the incentives for higher individual performance were now spread across all group members.56 - Clearly, as we saw in Chapter 7, pay can be a major factor in decisions to stay or leave. Data suggest that dissatisfaction with pay can be a key factor in turnover.57 Too little pay triggers feelings of being treated unfairly. The result? Turnover. Supporting this, pay that employees find reasonable can help reduce turnover.58 Even the way we pay has an impact on turnover. Evidence suggests that some employees are uncomfortable with pay systems that put any substantial future earnings at risk or pay systems that link less to personal effort and more to group effort.59 Another recent study found that superior-performing employees were less likely to leave if they received bonuses. No such positive result was found with pay increases (thus changing base pay) in that study.60 We need to make sure, as one critic has noted, that we don't let our design of new reward systems rupture our relationships with existing employees.61 However, as we have argued, negative sorting effects (high performers leaving) are more likely if base-pay growth does not keep up with performance over time.62 Recent efforts to use different types of compensation as a tool for retaining workers have focused on what is called scarce talent. For example, information technology employees have been scarce for much of the past decade, if not longer. One way to retain these workers is to develop a variable-pay component for each project. For example, reports of variable pay linked to individual length of stay on a project, to peer ratings, and to project results suggest that this pay-for-performance combination may appeal to scarce talent.63 - The next time you go into an Applebee's restaurant, think about how the company has historically used compensation to reduce turnover. In an industry where manager turnover hovers around 50 percent, Applebee's has been known to allow general managers to earn as much as $30,000 above base salary for hitting targets for sales, profitability, and customer satisfaction. To discourage turnover, this extra compensation is deferred for two years.64 - Besides money, other rewards also influence the decision to stay in a firm (retention). According to one recent study, the rewards that are effective in helping to retain employees in tough economic times are as follows:65 Do Employees More Readily Agree to Develop Job Skills Because of Pay? - We don't know the answer to this question. Skill-based pay (Chapter 6) is intended, at least partially, to pay employees for learning new skills—skills that hopefully will help employees perform better on current jobs and adjust more rapidly to demands on future jobs. For example, the U.S. Army pays ROTC cadets in college to learn new languages. Hot spots like the Mideast command monthly premiums of $100 to $250 per month.66 Anyone know Farsi (spoken in Iran)? - We do know that one complaint about skill-based pay centers on cost implications. More employees request training, spurred by the promise of skill-based increments. Poorly administered plans, allowing more people to acquire certification in a skill than are actually required, creates cost inefficiencies. This leads to plan abandonment. So is the net result positive? Whether the promise of skill-based pay is fulfilled is unclear. Evidence suggests that while pay for skill may sometimes but not always increase productivity, it does focus people on believing in the importance of quality and in turning out significantly higher quality products.67 Do Employees Perform Better on Their Jobs Because of Pay? - No matter what stand you take on this question, someone is going to disagree with you.68 However, a well-designed plan linking pay to behaviors of employees generally results in better individual and organizational performance.69 One study looked at the HR practices of over 3,000 companies.70 One set of questions asked: (1) Did the company have a formal appraisal process, (2) Was the appraisal tied to the size of pay increases, and (3) Did performance influence who would be promoted? Organizations significantly above the mean (by one standard deviation) on these and other "high-performance work practices" had annual sales that averaged $27,000 more per employee. WorldatWork surveyed 1,001 organizations (most, but not all, in the private sector) and 92 percent reported they used individual performance to determine salary increases. Further, 84 percent also used variable pay that was tied to performance.71 - In his review, Heneman reports that 40 of 42 studies looking at merit pay show higher performance when pay is tied to performance.72 Strong evidence suggests that linking pay to performance does increase motivation of workers and lead to improved performance.73 Locke and colleagues analyzed studies where individual incentives were introduced into actual work settings. Productivity increased on average 30 percent.74 Other meta-analyses draw similar conclusions—money does motivate performance.75 - Looking at this from the opposite direction, consulting firm data indicate that high performers received not only significantly higher merit increases than average performers (4.5 percent versus 2.6 percent) but also higher bonuses (140 percent of target versus 99 percent of target).76 One study of 841 union and nonunion companies found gain-sharing and profit-sharing plans (both designed to link pay to performance) increased individual and team performance 18 to 20 percent.77 How, though, does this translate into corporate performance? A review of 26 studies gives high marks to profit-sharing plans: Organizations with such plans had 3.5 to 5 percent higher annual performance.78 Gerhart and Milkovich took the performance-based pay question one step further. Across 200 companies they found an 8 to 20 percent increase in return on assets for a 10 percentage point (from 10 percent to 20 percent) increase in the size of a bonus (as a percentage of base salary).79 Further, they found that the variable portion of pay (which also included the percent of employees eligible for long-term incentives) had a stronger impact on individual and corporate performance than did the level of base pay. - If all this research isn't enough to convince you, consider Google's position on the matter. In recent speeches the top HR person at Google, Laszlo Bock, says the people you hire and the way you treat them (the rewards) make for happier and more productive workers. With evangelical zeal he advocates transparency in communications, obtainable goal setting, and less hierarchical organizations (fewer levels so employees can empower themselves to grow and learn).80 Bock also advocates "Pay unfairly (it's more fair!)." Bock explains that a small percentage of employees create a large percentage of the value and that their pay must recognize their disproportionate contributions. Conversely, numerous critics, led by Alfie Kohn, argue that incentives are both morally and practically wrong.81 The moral argument suggests that incentives are flawed because they involve one person controlling another. The counterargument to this notes that employment is a reciprocal arrangement. In periods of low unemployment especially, workers can choose whether they want to work under compensation systems with strong pay-for-performance linkages (as in the case of incentive systems). We do know that applicants aren't totally risk-averse. Some, especially high performers, tend to prefer performance-based pay rather than a totally fixed salary. Generally, if pay depends on individual performance, applicants find the company more attractive. Team-based incentives, in contrast, are less attractive, although it depends on the person. Kohn also suggests that incentive systems can actually harm productivity, a decidedly negative practical outcome. His rationale is based on citations mostly to laboratory studies where subjects work in isolation on a task for either pay or no pay. His conclusion, based heavily on the work of Deci and colleagues, is that rewarding a person for performing a task reduces interest in that task—extrinsic rewards (money) reduce intrinsic rewards (enjoyment of the task for its own sake).82 Critics of this interpretation point out at least two important flaws in Kohn's conclusions.83 First, the pragmatics of business demand that some jobs be performed—indeed, many jobs—that aren't the most intrinsically interesting. Although Target may be a great store for shopping, spending day after day stocking shelves with towels and other nonbreakables falls far down the intrinsic-interest scale.84 If incentives are required for real-world jobs to be completed and thus to create value for an organization and its consumers, so be it. This may simply be one of the costs of doing business. However, the idea that pay-for-performance would only be used to compensate people for doing uninteresting jobs flies in the face of reality. As we have seen, pay-for-performance is much more likely to be used for higher-level jobs and the payouts are also much higher in higher level jobs. That is likely because high level jobs have more impact on organization performance. Higher level jobs are also interesting, challenging (yes, intrinsically motivating), but people in intrinsically interesting jobs still want to be recognized for their achievements, earn a good living for themselves and their families, and be paid equitably (in line with other high achievers).85 Second, studies cited by Kohn frequently looked at people in isolation. In the real world people interact with each other, know who is performing and who isn't, and react to this when rewards are allocated. Without any link to performance, the less-motivated employees will eventually recognize that harder work isn't necessary. These issues raise the basic question, "Should we tie pay to performance?" One view says, "Not always." Employers are less likely to offer performance-based pay when the job involves multitasking, important quality control issues, or team work. In all three cases performance is harder to measure, and employers shy away from linking pay as a consequence.86 Alternatively, we could break the issue down to a series of questions. The first question perhaps should focus on an obvious but often overlooked issue: Do employees think any link at all should be made between pay and performance? Substantial evidence indicates that management and workers alike believe pay should be tied to performance. Dyer and colleagues asked 180 managers from 72 different companies to rate nine possible factors in terms of the importance they should receive in determining the size of salary increases.87 This group believed the most important factor for salary increases should be job performance. Following close behind is a factor that presumably would be picked up in job evaluation (nature of job) and a motivational variable (amount of effort expended). Other research supports these findings.88 Another way to make the pay-for-performance argument is to look at the ways HR professionals try to cut costs. At the top of the list: Create greater distinction between high and low performers!89 In other words, really pay for performance! Once we move away from the managerial ranks, though, other groups express a different view of the pay-performance link. The role that performance levels should assume in determining pay increases is less clear-cut for blue-collar workers.90 As an illustration, consider the frequent opposition to compensation plans that are based on performance ratings (merit pay). Unionized workers prefer seniority rather than performance as a basis for pay increases.91 Part of this preference may stem from a distrust of subjective performance measurement systems. Unions ask, "Can management be counted on to be fair?" In contrast, seniority is an objective index for calculating increases. Some evidence also suggests that women might prefer allocation methods not based on performance.92 It's probably a good thing that, in general, workers believe pay should be tied to performance, because the research we've reported suggests this link makes a difference.93 How does this performance improvement occur? One view suggests, as we have emphasized (see Chapter 1), that linking pay to performance occurs through two mechanisms, an incentive effect and a sorting effect.94 The incentive effect means pay can motivate current employees to perform better. The sorting effect means people sort themselves by what is important to them. So if Company X pays for performance, and you don't want to play by those rules (i.e., work harder or smarter to perform better) you sort yourself out, most easily by leaving Company X and finding another company with different rules for getting rewards. Recall the Lazear study we discussed in Chapter 1. Employee productivity (windshields installed per employee) at Safelite Glass increased 44 percent when pay practices switched from salaries (which were unlinked to windshields installed) to individual incentives (where pay depended on number of windshields installed). The interesting hook in this study was a separate analysis of employees who were present before and after the change in pay system (stayers) versus those who left and were replaced. Only one-half of the 44 percent increase (i.e., a 22 percent increase) came from increased productivity of the "stayers" (i.e., the incentive effect). The other half (i.e., the other 22 percent increase) was due to the sorting effect: Less productive workers (who did worse under the new incentive system) left and were replaced by more productive workers who would earn more money under the new incentive system.95 Thus, people sort themselves into or out of organizations based on a preference for being paid based on personal performance or something else.96 Of course, the most obvious sorting factor is ability. Higher-ability individuals are attracted to companies that will pay for performance, thus recognizing their greater contribution.97 High performers will also leave firms that don't reward their performance (pay for something like seniority rather than performance) and go to those that do. Coming at this from a different angle, data also suggest that some people make job choices based on effort aversion. "Find me a job where I don't have to work hard!" Obviously, jobs where high performance is expected are less likely to be attractive to these prospective employees.98 When we look at pay and group performance (instead of individual performance), the evidence is mixed. In general, though, we think that group pay (whether the group is a team or an entire organization) leads to small increases (relative to individual pay for performance) in productivity. A recent study suggests that, to be effective, group incentives are most effective when paired with complementary HR practices. Specifically, group incentives work if you have implemented a team-based structure where members monitor team performance and personally sanction "free riders."99 Before we rush to add a variable-pay component (one form of pay for performance) to the compensation package, though, we should recognize that such plans can, and do, fail. Sometimes, ironically, the failure arises because the incentive works too well, leading employees to exhibit rewarded behaviors to the exclusion of other desired behaviors. (See Chapter 10 for further discussion of the risks of such plans.) Exhibit 9.7 documents one such embarrassing incident that haunted Sears for much of the early 1990s.100 Apparently the Sears example is no fluke. Other companies have found that poorly implemented incentive pay plans can hurt rather than help. Green Giant, for example, used to pay a bonus based on insect parts screened in its pea-packing process. The goal, of course, was to cut the number of insect parts making their way into the final product (anyone planning on vegetables for dinner tonight?). Employees found a way to make this incentive system work for them. By bringing insect parts from home, inserting, and inspecting, their incentive dollars rose. Clearly, the program didn't work as intended. Experts contend this is evidence that the process wasn't managed well. The second author of your book also recollects the first author was working with an oil drilling company. They attached a sizable incentive to compensation for geologists if they could reduce the time to explore for a well site and set up the drilling operation. They couldn't figure out why so many more wells were being reported as dry (no oil) than had been the case historically!101 (Could it be that well sites were suddenly being chosen more on the basis of their ease of setting up a drilling operation—perhaps those in less remote locations—and chosen less on the basis of how much oil is in the ground there?) What does this mean in terms of design? We return to the risks (as well as the potential returns) possible from the use of incentive pay in Chapter 10.

DESIGNING A PAY-FOR-PERFORMANCE PLAN

Our pay model suggests effectiveness is dependent on three things: efficiency, equity, and compliance in designing a pay system. Efficiency= Efficiency involves three general areas of concern. - Strategy: Does the pay-for-performance plan support corporate objectives? For example, is the plan cost-effective, or are we making payouts that bear no relation to improved performance on the bottom line? Similarly, does the plan help us improve quality of service? Some pay-for-performance plans are so focused on quantity of performance as a measure that we forget about quality. Defect rates rise. Customers must search for someone to handle a merchandise return. A number of things happen that aren't consistent with the emphasis on quality that top organizations insist upon. The plan also should link well with HR strategy and objectives. If other elements of our total HR plan are geared to select, reinforce, and nurture risk-taking behavior, we don't want a compensation component that rewards the status quo. Be careful, though. The Federal National Mortgage Association, also known as Fannie Mae, which buys mortgages from banks and re-sells them to investors, changed its performance metrics from return on assets and cost management to total earnings and earnings per share. No problem, you say? One view is that it was a problem because the CEO of Fannie Mae was part of the "team" that wrote legislation and he took this as an opportunity to influence the laws to fit his company's strategy. In this telling, he got rich but also helped start the 2008 financial crisis.102 Finally, we address the most difficult question of all—how much of an increase makes a difference? What does it take to motivate an employee? Is 3 percent, the recent average of pay increases, really enough to motivate higher performance? One review of the evidence suggests that an increase must be at least 6 to 7 percent "to be seen as meaningful" and goes on to say: "Obviously, people prefer any raise to no raise at all. But small raises, when presented as rewards for merit, can be dysfunctional. Organizations with small pay raise pools may wish to think seriously about their allocation of merit raises."103 - Structure: Is the structure of the organization sufficiently decentralized to allow different operating units to create flexible variations on a general pay-for-performance plan? For example, IBM adapted performance reviews to the different needs of different units, and the managers in them, resulting in a very flexible system. In this system, midpoints for pay grades don't exist. Managers get a budget, some training on how to conduct reviews, and a philosophical mandate: Differentiate pay for stars relative to average performers, or risk losing stars. Managers are given a number of performance dimensions. Determining which dimensions to use for which employees is totally a personal decision. Indeed, managers who don't like reviews at all can input merit increases directly, anchored only by a brief explanation for the reason.104 Different operating units may have different competencies and different competitive advantages. We don't want a rigid pay-for-performance system that detracts from these advantages, all in the name of consistency across divisions. - Standards: Operationally, the key to designing a pay-for-performance system rests on standards. Specifically, we need to be concerned about the following: * Objectives: Are they specific yet flexible? Can employees see that their behavior influences their ability to achieve objectives (called the "line-of-sight" issue in industry)? * Measures: Do employees know what measures (individual appraisals, peer reviews of team performance, corporate financial measures, etc.) will be used to assess whether performance is sufficiently good to merit a payout? * Eligibility: How far down the organization will the plan run? Companies like PepsiCo and Starbucks believe all employees should be included. Others think only top management can see how their decisions affect the bottom line. * Funding: Will you fund the program out of extra revenue generated above and beyond some preset standard? If so, what happens in a bad year? Many employees become disillusioned when they feel they have worked harder but economic conditions or poor management decisions conspire to cut or eliminate bonuses. Equity/Fairness - Our second design objective is to ensure that the system is fair to employees. Two types of fairness are concerns for employees. The first type is fairness in the amount that is distributed to employees. Not surprisingly, this type of fairness is labeled distributive justice.105 Does an employee view the amount of compensation received as fair? As we discussed earlier in the section on equity theory, perceptions of fairness here depend on the amount of compensation actually received relative to input (e.g., productivity) compared against some relevant standard. Notice that several of the components of this equity equation are frustratingly removed from the control of the typical supervisor or manager working with employees. A manager has little influence over the size of an employee's paycheck. It is influenced more by external market conditions, pay-policy decisions of the organization, and the occupational choice made by the employee. Indeed, recent research suggests that employees may look at the relative distribution of pay. For example, some major league baseball teams have met with mixed success in trying to buy stars via the free-agent market. Some speculate that this creates feelings of inequity among other players. Some evidence suggests that narrower ranges for pay differences may actually have positive impacts on overall organizational performance.106 - Managers have somewhat more control over the second type of equity. Employees are also concerned about the fairness of the procedures used to determine the amount of rewards they receive. Employees expect procedural justice.107 Evidence suggests that organizations using fair procedures and having supervisors who are viewed as fair in the means they use to allocate rewards are perceived as more trustworthy and command higher levels of commitment.108 Some research even suggests that employee satisfaction with pay may depend more on the procedures used to determine pay than on the actual level distributed.109 - A key element in fairness is communications. Employees want to know in advance what is expected of them. They want the opportunity to provide input into the standards or expectations. And if performance is judged lacking relative to these standards, they want an appeals mechanism. The importance of open communication also extends to upper management. In firms that practice greater transparency of pay practices, executives perform better when told what the linkage is between pay and performance.110 In a union environment, this is the grievance procedure. Something similar needs to be set up in a nonunion environment.111 As evidence, only 15 percent of employees who feel well informed indicate they are considering leaving their company. This jumps to 41 percent who think about leaving if they feel poorly informed about the way the pay system operates.112 Compliance - Finally, our pay-for-performance system should comply with existing laws. We want a reward system that maintains and enhances the reputation of our firm. Think about the companies that visit a college campus. The interview schedules fill up quickly for some of these companies because students gravitate to them. Why? The companies' reputations.113 We tend to undervalue the reward value of a good reputation. To guard this reputation, we need to make sure we comply with compensation laws.

WHAT BEHAVIORS DO EMPLOYERS CARE ABOUT? LINKING ORGANIZATION STRATEGY TO COMPENSATION AND PERFORMANCE MANAGEMENT

The simple answer is that employers want employees to perform in ways that lead to better organizational performance. Exhibit 9.1 shows how organizational strategy is the guiding force that determines what kinds of employee behaviors are needed. As an illustration, Nordstrom's department stores are known for extremely good quality merchandise and high levels of customer satisfaction—this is the organization strategy they use to differentiate themselves from competitors. Nordstrom's success isn't a fluke. You can bet that some of their corporate goals, strategic business unit goals (SBU goals, where a strategic business unit might be a store), department-level goals, and, indeed, individual employee goals are linked to pleasing customers and selling high-quality products. The job of Human Resources is to devise policies and practices (and compensation falls in this mix) that lead employees (the last box in Exhibit 9.1) to behave in ways that ultimately support corporate goals. When you walk into a Nordstrom's, you see employees politely greeting you, helping without suffocating, and generally making the shopping experience a pleasant one. These are behaviors that support Nordstrom's strategic plan. Every organization, whether they realize it or not, has human resource practices that can either work together or conflict with each other in trying to generate positive employee behaviors. One way of looking at this process is evident from Exhibit 9.2, which says that behavior is a function of the ability, motivation, and opportunity to perform. We first saw this idea in Chapter 2. Let's use an example from the classroom. You have to make a presentation tomorrow. Twenty percent of your grade (the reward) depends on it. Do you have the ability? Can you speak clearly, be interesting, and send a good message? Are you motivated? (Or is this class unimportant to you? Would you prefer to watch the finale of "Walking Dead" on TV?) And is the opportunity to do well (e.g., without environmental obstacles) present? (One of your authors was making a presentation a few years ago in Indonesia when an earthquake hit. Ruined the speech!) Wanting to succeed isn't enough. Having the ability but not the motivation also isn't enough. Many a player with lots of talent doesn't have the motivation to endure thousands of hours of repetitive drills, or to endure weight training and general physical conditioning. Even with both ability and motivation, a player's work environment (both physical and political) must be free of obstacles. A home run hitter drafted by a team with an enormous ball park (home run fences set back much farther from home plate) might never have the opportunity to reach his full potential. The same thing is true in more traditional jobs. Success depends on finding people with ability—that's the primary job of recruitment, selection, and training. Once good people are hired, they need to be motivated to behave in ways that help the organization. (Note, part of selection is also to hire motivated people, so the triangles interact with each other, as denoted by the three-pronged arrow in the center of Exhibit 9.2.) This is where compensation enters the picture. Pay and other rewards should reinforce desired behaviors. But so, too, should performance management, by making sure that what is expected of employees, and what is measured in regular performance reviews, is consistent with what the compensation practices are doing. And perhaps most important of all, the culture of the organization (the informal rules and expectations that are evident in any company) should point in the same direction. Finally, HR needs to establish policies and practices that minimize the chances that outside "distractors" hinder performance. In the 1980s, Nabisco was slow to recognize customer demand for "soft batch" cookies. Why? They had a centralized organization structure in which it took a long time for sales information to reach the top decision makers. No matter how much ability or motivation the sales staff has, it's hard to sell cookies the public doesn't want. What did Nabisco do? They decentralized the company (organization design), creating divisions responsible for different product lines. Now when sales people say consumer preferences are changing, response is much more rapid. Similarly, if we don't recognize changing skill requirements (human resource planning), it's hard to set up revised training programs or develop compensation packages to reward these new skills instantly. Knowing in advance about needed changes makes timely completion easier. As a further illustration, if we have inefficient processes (such as too many steps in getting approval for change), organization development (the process for changing the way a company operates) can free up motivated workers to use their skills. The key lesson from Exhibit 9.2 is important: Compensation can't do it all alone. Try changing behavior by developing a compensation system to reward/motivate that behavior. If you haven't selected the right people (and/or provided the necessary training/development) or if you haven't designed the work and the culture to provide the opportunity to leverage employees' motivation and ability, you're destined for failure. Not taking into account the importance of differences in how people respond to different pay systems is another potential pitfall.5 When you're out working in HR, and your boss asks you to "fix" something (such as pay or performance appraisal), make sure the change fits with what other HR programs in the organization are trying to do. Otherwise trouble lurks. So, what behaviors does compensation need to reinforce? First, compensation should be sufficiently attractive to make it possible to recruit and hire good potential employees (attraction).6 Second, we need to make sure the good employees stay with the company (retention). The recession of 2008-2010, which still has a lingering impact today, severely tested companies' layoff strategies. Losing money? Layoff workers! But sometimes laying off workers means cutting future stars. Some wiser companies kept costs down by reducing labor costs without layoffs—by, for example, temporarily reducing salaries, bonuses, 401(k) retirement contributions, and/or work hours. Caterpillar, FedEx, Black and Decker, Honeywell, the New York Times, and the State of Pennsylvania are examples of organizations that reduced labor costs without layoffs.7 If we can succeed at these first two things, we can then concentrate on building further knowledge and skills (develop skills). And, finally, we need to find ways to motivate employees to perform well in their jobs—to take their knowledge and abilities and apply them in ways that contribute to organizational performance. For example, whether we've developed a sound compensation package can only be determined by the impact on performance. We can't tell if our compensation system helps recruit and select good employees if we don't know how to measure what constitutes good. We can't tell if employees are building the kinds of knowledge base they need if we can't measure knowledge accumulation. We can't reward performance if we can't measure it! As a simple example, think about companies where piece-rate systems are used to pay people. Recently, one of the authors went to a sawmill, looking for wood to build a dining room table. Talks with the manager revealed they used a simple piece-rate system to motivate workers. For every board foot of lumber cut using giant saws, workers were paid a fixed amount. There is little ambiguity in this measure of performance, and this makes it easy to create a strong link between units of performance and amount of compensation. One of the biggest recent advances in compensation strategy has been to document and extend this link between ease of measuring performance and the type of compensation system that works best. Let's take a minute to talk about each of the cells in Exhibit 9.3. They help explain why incentives work in some situations and not in others. The columns in Exhibit 9.3 divide companies into those with widely variable performance from year to year and those with much more stable performance across time. What might cause wide swings in corporate performance? Often this occurs when something in the corporation's external environment fluctuates widely. A perfect example is gas prices. Sometimes when airlines are reporting high profits, it is not because they've done anything particularly well, but because airline fuel costs have dropped significantly.8 It probably wouldn't be fair, and employees would certainly object, if a large part of pay were incentive-based in this kind of environment. Lack of employee control translates into perceptions of lack of opportunity to perform well and unfair treatment if pay is tied to these uncontrollable things. Situations B and D both suggest that a low-incentive component is appropriate in organizations with highly variable annual performance. Conversely, as situations A and C indicate, larger-incentive components are appropriate in companies with stable annual performance. The rows in Exhibit 9.3 note that individual employee performance also can vary. Some jobs are fairly stable, with expectations fairly consistent across time. What I do today is basically the same thing I did yesterday. And tomorrow looks like a repeat too! In other jobs, though, there might be high fluctuation in the kinds of things expected of employees, and for these jobs there is high demand for employees who are willing to be flexible and adjust to changing demand. Here, using incentive pay exclusively might not work. Incentive systems are notorious for getting people to do exactly what is being incentivized. Pay me big money to sell suits, and that's just what I'm going to do. You want me to handle customer returns, too? No way, not unless the compensation system rewards a broader array of duties. Evidence suggests that companies are best able to get employees to adjust, be flexible, and show commitment when a broader array of rewards, rather than just money, is part of the compensation package.9 For example, why does Lincoln Electric (a major producer of welding machines) outproduce other companies in the same industry year after year? Normally we think it's because the company has a well-designed incentive system that links to level of production. Certainly this is a big factor! But when you talk to people at Lincoln Electric, they suggest that part of the success comes from other forms of reward, including the strong commitment to job security—downsizing simply isn't part of the vocabulary there—which reinforces a willingness to try new technologies and new work processes (a culture that supports innovation). Situation A, with low variability in corporate performance but unclear performance measures for employees, describes the kind of reward package that fits these job and organizational performance characteristics. When we distill all of this, what can we conclude? We think the answer depends on how we respond to the following four questions: 1. How do we attract good employment prospects to join our company? 2. How do we retain these good employees once they join? 3. How do we get employees to develop skills for current and future jobs? 4. How do we get employees to perform well while they are here? First, how do we get good people to join our company? How did Nike get LeBron James to serve as a corporate spokesperson? Part of the answer is cold hard cash—$90 million guaranteed before LeBron had even played his first NBA game.10 But it isn't always just about the money. LeBron asked for and got artistic input on the products designed by Nike. Even when the decision doesn't involve millions of dollars, the long-run success of any company depends on getting good people to accept employment. One particularly good study of this very question found that job characteristics (including rewards and tasks/abilities required) and recruiter behaviors were key elements in the decision to join a company.11 Second, the obvious complement to the decision to join is the decision to stay. How do we retain employees? It doesn't do much good to attract exceptional employees to our company only to lose them a short time later. Once our compensation practices get a good employee in the door, we need to figure out ways to ensure it's not a revolving door. Max Scherzer left the Detroit Tigers, a perennial pick for a division title, to play with the Washington Nationals. Was it the $210 million contract, including a $50 million signing bonus and an annual average salary of $30 million? Or was it because the Washington Nationals are favored to win the World Series? What does it take to retain key people? Money?12 Or are other rewards important? And does their absence lead us to use money as the great neutralizer? Third, we also must recognize that what we need employees to do today may change—literally overnight! A fast-changing world requires employees who can adjust more quickly. How do we get employees, who traditionally are resistant to change, to willingly develop skills that might not be vital on the current job but are forecast to be critical in the future? Another compensation challenge! Finally, we want employees to do well on their current jobs. This means performing—and performing well—tasks that support our strategic objectives. What motivates employees to succeed? The compensation challenge is to design rewards that enhance job performance.

The primary focus of Part 3 was on determining the worth of jobs, independent of who performed those jobs. Job analysis, job evaluation, and job pricing all have a common theme. They are techniques to identify the value a firm places on its jobs. Now we introduce people into the equation. Now we declare that different people performing the same job may add different value to the organization. Wesley is a better programmer than Kelly. Erinn knows more programming languages than Ian. Who should get what?

If you stop and think about it, employment is about a contract. When you accept a job, you agree to perform work—to complete tasks—in exchange for rewards. Sometimes this contract is simple. You agree to cut your neighbor's grass for $25. When you're done, she pays you the money. As we grow up, the contracts become more complex. Your first job after college will involve a contract, although the terms of that contract won't be spelled out fully. You will sign an agreement to perform tasks to the satisfaction of your employer. Rarely are the "satisfaction" terms clearly defined upfront. You hope you'll get feedback along the way that helps make the performance standards more apparent. In exchange, the contract reads, you will receive a certain amount of pay. Let's hope it's a big number! But you might be a bit disappointed in the rest of the contract, because it's rare that there will be any mention of other rewards (e.g., how interesting your work is and how collaborative your co-workers are) besides money (and perhaps related benefits). You take it on faith that the employer will do right by you beyond just the monetary component of the job. At the extreme of employment contracts are those between an employer and a union representing workers (Chapter 15). Unionized employees aren't willing to take it on faith that the employer will provide fair compensation for work performed. Some union contracts are hundreds of pages long and specify in the tiniest detail what a worker will get as rewards under different work conditions. Whether the employment contract is simple or complex, though, we still struggle with both sides of the agreement. How do we decide if a worker is performing the job satisfactorily and how much should we pay for that performance? Entering people into the compensation equation greatly complicates the compensation process. People don't behave like robots. Believe us, the auto industry has tried replacing people with robots. On some jobs, like welding car parts together, robots work just fine. Robots can tighten a bolt and oil a joint. But for many jobs (though not as many as in the past) it's easier and cheaper to do things the old-fashioned way—with people. The challenge is to design a performance and reward system so that employees support what the company is trying to accomplish. Indeed, there is growing evidence that the way we design HR practices, like performance management, strongly affects the way employees perceive the company. Well-known organizations, like The Container Store and Costco, both pay more and have more sophisticated performance-tracking methods that lead to happier employees.1 And this directly affects corporate performance.2 The simple (or not so simple, as we will discuss) process of implementing a performance appraisal system that employees find acceptable goes a long way toward increasing trust for top management.3 And that new performance appraisal and reward system has an impact on other parts of HR. The pool of people we recruit and select from changes as our HR system changes. In Chapter 2, we talked about sorting effects. Not everyone "appreciates" an incentive system or even a merit-based pay system. People who prefer pay systems that are less performance-based will "sort themselves" out of organizations that have these pay practices and philosophies. Those people either won't respond to recruitment ads, or, if already employed, may seek employment elsewhere.4 So as we discuss pay and performance in this chapter and in Chapters 10 and 11, remember that there are other important outcomes that also depend on building good performance measurement tools. In Chapter 1 we talked about compensation objectives complementing overall human resource objectives and both of these helping an organization achieve its overall strategic objectives. But how does an organization achieve its overall strategic objectives? In this part of the book, we argue that organizational success ultimately depends on human behavior. Why did the St. Louis Cardinals in 1967, the Cleveland Cavaliers in 2016, and the Philadelphia Eagles in 2018 win a championship? Answer: It was a team effort, but it sure helped to have had on those teams, respectively, Bob Gibson ("Rapid Robert"), LeBron James ("Greatest player on the planet"), and Nick Foles (no real nickname just yet)—for example, Bob Gibson went 3-0 (in a best-of-seven series!) and gave up a total of three earned runs in those three (complete) games he pitched. (The New England Patriots won in 2019 for a change. Are Tom Brady and Bill Belichik still there?) Or how did Dumbledore's Army and the Order of the Phoenix win the Battle of Hogwarts? Surely, it was because of superior wizarding talent (and not being evil)! That's behavior in its simplest and purest form. Our compensation decisions and practices should be designed to increase the likelihood that employees will behave in ways that help the organization achieve its strategic objectives (incentive effect) and to attract and retain such employees to the organization (sorting effect). This chapter is organized around employee behaviors. First, we identify the four kinds of behaviors organizations are interested in. Then we note what theories say about our ability to motivate these behaviors. And, finally, we talk about our success, and sometimes lack thereof, in designing compensation systems to elicit these behaviors.

WHAT DOES IT TAKE TO GET THESE BEHAVIORS? WHAT PRACTITIONERS SAY

In the past, compensation people didn't ask this question very often. Employees learned what behaviors were important as part of the socialization process or as part of the performance management process.24 If it was part of the culture to work long hours, you quickly learned this. One of our daughters worked as a business consultant for Accenture, a very large consulting company. She learned quickly that 70- to 80-hour work weeks were fairly common. Sure, she had very good wages for someone with a bachelor's degree in biology and no prior business experience, but it didn't take long to burn out when weeks of long hours turned into months. If your performance appraisal at the end of the year stressed certain types of behaviors, or if your boss said certain things were important to her, then the signals were pretty clear: Do these things! Compensation might have rewarded people for meeting these expectations, but usually the compensation package wasn't designed to be one of the signals about expected performance. That has not been true for many years!25 Now compensation people talk about pay in terms of a neon arrow flashing "Do these things." Progressive companies ask, "What do we want our compensation package to do? How, for example, do we get our product engineers to take more risks?" Compensation is then designed to support this risk-taking behavior. Compensation people will also tell you, though, that money isn't everything. But they will also tell you that money (and the sense of achievement and recognition that go with it) is very important. Indeed, there is no better way to ensure that you will hear from an employee than if that employee feels his/her pay is not equitable/fair, given what that employees sees as his/her contributions. e-compensation - The International Society for Performance Improvement has web information on performance journals, strategies for improving performance, and conferences covering the latest research on performance improvement techniques. Go to the society's website, www.ispi.org. Here, however, we do want to identify the many other rewards in addition to compensation that influence employee behavior. Sometimes this important point is missed by compensation experts. Going back at least to Henry Ford, we tend to look at money as the great equalizer. Job boring? No room for advancement? Throw money at the problem! More money is a solution sometimes (e.g., as a compensating wage differential to offset undesirable job attributes). In other cases, however, workers may value improvement in another reward area. To find out, you can ask them (e.g., in a survey). Surveys of workers in general (not as useful as a survey of your own workforce, but still of some use) report that workers generally also highly value other job rewards such as empowerment, recognition, and opportunities for advancement.26 Entire books, for example, illustrate hundreds of ways to give meaningful recognition to employees.27 And there is growing sentiment for letting workers choose their own "blend" of rewards from the 13 we note in Exhibit 9.5. We may be overpaying in cash and missing the opportunity to let employees construct both a more satisfying and less-expensive reward package. Known as flexible compensation, this idea introduced earlier is based on the notion of different rewards having different dollar costs associated with them. Armed with a fixed sum of money, employees move down the line, buying more or less of the 13 rewards as their needs dictate.28 While widespread use of this type of system may be a long time in the future, the cafeteria approach still underscores the need for integration of rewards in compensation design. If we don't think about the presence or absence of rewards other than money in an organization, we may find the compensation process producing unintended consequences. Consider the following three examples, which show how compensation decisions have to be integrated with total reward system decisions: Example 1: A few years ago McDonald's completed a worldwide "employment branding" exercise. Their goal was to find out what people liked about jobs at McDonald's and to feature these rewards in the recruitment strategy for new employees. Three things emerged as strengths at McDonald's: (1) an emphasis on family and friends in a social work environment; (2) flexibility in work assignments and work schedules; and (3) development of skills that helped launch future careers.29 Example 2: This example comes from airline industry leader Southwest Airlines.30 Southwest Airlines promotes a business culture of fun and encourages employees to find ways to make their jobs more interesting and relevant to them personally. All this is accomplished (at least on the surface) without using incentives as a major source of competitive advantage. Indeed, pay at Southwest isn't any higher than for competitor airlines, yet it's much easier to recruit top people there. Fun, a good social environment, is a reward! A somewhat different view is that fun and a good social environment are indeed crucial to Southwest's culture and competitive advantage. However, it may be that incentives are too. Southwest "pays by the trip," meaning that employees earn more, the more flights they work. (Planes only make money for Southwest when they are in the air.) Southwest employees also have significant ownership in the company through a discount stock purchase program, and they receive profit-sharing checks when the airline is profitable, which it has been for 45 years in a row!31 For example, in the most recent year, Southwest paid out $543 million in profit sharing to its 56,110 employees, which works out to an average of $9,677 per employee.32 Of course, earning more money for working more flights and from ownership and profit sharing only happens when Southwest does well. That is an example of alignment between the interests of Southwest Airlines and its employees. Add fun and a good social environment, and apparently you can be profitable 45 years in a row, even in an industry marked by regular bankruptcies. Example 3: Consider the relationship between the different forms of compensation and another of the general rewards listed in Exhibit 9.5: security. Normally we think of security in terms of job security. Drastic reductions in middle-management layers during the downsizing decade of the 1980s increased employee concerns about job security and probably elevated the importance of this reward to employees today. Maybe that's why new millennial workers are concerned not only about employment risk but also about compensation risk. There is evidence that compensation at risk (variable pay, which is a payment based on performance objective achievement that does not become part of base pay in future years) may leave employees less satisfied both with their pay level and with the process used to determine pay.33 Security as an issue, it appears, is creeping into the domain of compensation. It used to be fairly well established that employees would make more this year than they did last year, and employees counted on such security to plan their purchases and other economic decisions. The trend today is toward less stable and less secure compensation packages. The very design of compensation systems today contributes to instability and insecurity. And that in turn leads some potential employees to reject a firm or some current employees to decide it's time to leave. Of course, it may be the case that when such plans offer enough upside earnings potential for high performance that high performers may actually prefer to work under such plans (see incentive and sorting effects discussed in Chapter 1, later in this chapter, and in Chapter 10). Exhibit 9.6 outlines the different types of compensation components, roughly in order from least risky to most risky for employees. We define risk as variance (lack of stability) of income and/or the inability to predict income level from year to year. Base pay is, at least as far as there are any guarantees, the guaranteed portion of income, as long as employees remain employed. Since the Depression there have been very few years when base wages did not rise or at least stay the same.34 Across-the-board increases, cost-of-living increases, and merit increases all help the base pay component increase on a regular basis. Of course, there always has to be an exception to the rule—and the Great Recession of 2008-2010 spawned many cuts in corporate base wages. The next seven components are distinguished by increasing levels of risk/uncertainty for employees. In fact, risk-sharing plans actually include a provision for cuts in base pay that are only recaptured in years when the organization meets performance objectives. For example, under a plan at E.I. du Pont de Nemours & Company, all employees put 6 percent of their salaries at risk and were paid a sliding percent of this based on how close the firm comes to its annual goals. If the company achieved less than 80 percent of goal, there was no pay increase. In the 80 to 100 range the increase was 3 to 6 percent. At 101 to 150 percent, increases were very lucrative to "pay for" employees taking a risk in the first place—in the 7 to 19 percent range.35 All of this discussion of risk is only an exercise in intellectual gymnastics unless we add one further observation: Over the last several decades, companies have been moving more toward greater use of variable pay programs, which are higher on the risk continuum. We will say more on the degree to which organizations use pay for performance, including variable pay, in Chapter 10. This greater compensation risk that employees are now asked to bear reflects the result of an evolution over time in forms of pay that are less entitlement-oriented and more variable and linked to individual, group, and corporate performance.36 Employees increasingly are expected to bear a share of the risks that businesses have solely born in the past. It's not entirely clear what impact this shifting of risk will have in the long run. Some are concerned that efforts to build employee loyalty and commitment may be a casualty of this greater use of variable pay and risk in pay systems.37 Indeed, some surveys suggest a downward trend in engagement levels for employees, although it may be that such trends, whether upward or downward, depend as much or more on the cyclical nature of the economy.38 Consistent with the idea of a compensating pay differential for risk, some surveys suggest that employees may need a risk premium (higher pay) to stay and perform in a company with pay at risk.39 Even a premium might not work for employees who are particularly risk-averse. Security-driven employees actually might accept lower wages if they come in a package that is more stable.40 On the other hand, some employees may be willing to accept higher risk in return for a greater chance to earn large payouts when performance is high. Also, some employees are not interested in a traditional employment model of lifetime employment with one organization and steady but modest merit increases over the course of their careers. As always, there are pros and cons to each approach, and employees will to some degree self-select/sort themselves into organizations that fit their preferences and out of organizations that do not. In any case, for the remainder of this chapter, we will return to the more general question of what impact these different forms of pay have on motivating the four general behaviors we noted earlier. e-compensation - IOMA is the Institute of Management and Administration. It specializes in finding studies from a wide variety of places that discuss different aspects of pay for performance. The index for IOMA's website is at www.ioma.com.