Personnel Economics

Personnel Economics Sorting and Investing in Employees Topic 1. Setting Hiring Standards - Economic concepts: - Asymmetric information: the two stakeholders don’t have the same amount of information (hiring process). The future employee or job prospect know his productivity in the given job but the recruiters don’t. - Investment in skills: sacri ce short-term potential in order to increase it in the future (education). - Economic transactions: the employee gives labour to the employer, who, in return, gives a remuneration. Other types of transactions are exchange between companies. - Employees: innate abilities (intelligence, creativity) and learned skills (education, experience, job training). 1.1 Hiring risky workers Example A company has 2 new applicants (G and S) and has to choose one of them. The method used to measure the more suitable worker for the company is calculating the expected output of the worker to the company E(x). Applicant G generates 200000€ of output, while applicant S has 50% probability of being a star worker that generates 500000€ of output; and 50% of being a total failure on his job, causing a loss of output of -100000€. E(x)G = 200000 € E(x)S = 0.5 · 500000€ + 0.5 (-100000€) = 200000€ - On average, it his expected that both applicants are equally productive E(x)G = E(x)S. This arises a problem because the company has to choose between a safe worker that knows is going to be productive or a risky worker that could be more or less productive than the other one. - Assumptions: - Both, S and G can be expected to work at the rm/company for 10 years. - The salary would be 100000€ annually, which would not have discounting value. - It takes 1 year to determine whether applicant S is a star employee or a very unproductive worker. - Firing is always possible without additional costs. - The company is approximately risk neutral. - Decision: E(π)G = 10 · (200000€ - 100000€) = 1000000€ E(π)S = 10 · (0.5 · (500000€ - 100000€)) + 1 · (0.5 · (-100000€ -100000€)) = 1900000€ retain re after rst year - Taking this into account, it makes sense to give it a try and hire the risky applicant. - Analysis: important factors in decision making - Downside Risk: more potential to destroy value. Less likely to be optimal hiring a risky worker. - Upside Potential: greater potential pro ts. Much to gain but little to loose. - Termination Costs: cost to re a worker. Higher expected cost of hiring a risky worker. - Risk Aversion: preference of not taking the risk. Risky workers are costly (higher value on avoiding the risk). - Lenght of Evaluation: the longer the time of evaluation, the less valuable to higher a risky candidate. - Lenght of Employment: the longer the worker stays in the rm, the larger the value of hiring a risky worker. - Counterargument: the assumptions do not apply. Pro t with a star employee: paying a low wage of 100000€ per year no matter how high-productive the star worker is. But, the rm might not be able to continue paying a star worker a low wage after he turns out to be a star. Cannot take advantage of its productivity anymore. fi fi fi fi fi fi fi fi fi 1 - Under which assumptions will there be a pro t from hiring a risky worker? - Asymmetric information between rms: the rms can pro t from their star workers just as long as they keep informational advantage about the star’s productivity. If a company has informational advantage about the stars productivity, can pro t from him. If other companies learn that this worker drives the performance of the company, might approach the worker and try to take him away from that company. Companies learn more about risky workers productivity during probation, and overtime in general. - Firm-speci c productivity: the productivity that a person display in a company doesn’t necessarily mean that it would be the same productivity that could display in other companies. - Immobility of the employee (multi period contracts): companies design contracts in a way that people have an incentive to stay in the company after the probation period. - Economic modeling: simpli cation of complex models. In order to make analysis, this simpli cation of problems is needed, and helps to make a decision. In analyzing risky hires, this economic concepts and tools are used: - Labor market competition for employees (o er of a hire wage from other companies). - Price (salary) - Asymmetric information: not knowledge about the pro tability of the applicant, and asymmetric information between companies. - Incentives: in order to keep star workers or in uence people to do something. 1.2 Balancing Bene ts Against Costs Example A company wants to hire enough workers to have 1000000€ in monthly revenue. There are 2 di erent types of applicants: HS, which are high school graduates employees, and Col, which are college graduates employees. The HS generate a monthly sales average of 122917€, while the Col generate 156944€. As productivity is here related with the salary, the monthly wage of HS is 2198€ per month and the wage of Col is 3455€ monthly.. πHS = 122917€ - 2198€ = 120719€ πCol = 156944€ - 3455€ = 153489€ - On average, the monthly pro t from Col, collage graduates, is more than the one the company would receive with the high school graduates. Hence to this, it is more pro table to higher Col applicants. - Assumptions: - Firms objective is to maximize the pro t - There are no hiring constraints - Prices for the output Q (productivity) and w (wage) are constant - Decision: Revenue / Monthly Sales = Number of Persons HS = 1000000€ / 122917€ = 8.1 Col = 1000000€ / 156944€ = 6.4 Number of Persons · Salary = Cost per Unit HS = 8.1 · 2198€ = 17804€ Col = 6.4 · 3455€ = 22112€ - Employing HS is more pro table because the cost per unit of output is lower. - Theorem in terms of wage w and output Q: - The most cost-e ective employee is the one that has the lowest salary to productivity ratio. ff fi fi fi fi fi fl ff fi fi fi fi fi fi fi fi ff fi 2 - Economic principles: - Tradeo between costs and bene ts (consider the ratios) - Compare the approach to the best alternative (opportunity costs) - Foreign competition: low labor cost countries drive companies in high labor costs countries out of business. Companies that establish many production plans in low labor cost countries in order to increase their pro t. - The real issue is not whether labor is cheaper, but whether it is more cost e ective. If a company moves the production to a country with low labor costs but with really ine cient and highly unproductive, the company might move the production again. - Low cost per unit of output. 1.3 The Production Function - Key decision in the short run: how much labor to be hired. Labor is heterogeneous factor of production (there are di erent types of workers). Standard economic theory of production: rms minimize total labor costs TC subject to production function. Variables: wages (w), output/productivity (Q), labor (L), capital (K), high school graduate labor (H) and college graduate labor (C). Total labor costs (TC). - Minimization Problem of the rm: - First-order-conditions (derivatives) with respect to H, C, λ: - Where the Lagrange multiplier λ re ects the marginal cost for a given Q. - λ can be determined using the rst-order-conditions - Q (optimal output) can be determined by setting the marginal cost (MC) equal to marginal returns (MR). - Key condition for the optimal skill ratio (independent of the output). Independently of how much output is want to be produced, there would be an optimal skill ratio. - The fraction between the rst-oder-conditions wH and wC: fi ff ffi ff fi fl fi fi fi fi ff 3 1. Workers productivity is independent of coworkers - Example: salespersons operating independently (responsible of di erent regions). The only thing that determines the performance of a salesperson is the business he does on the region he’s responsible of. Production factors 0 0; C = 0 or H = 0; C > 0. 2. Workers productivity depends on coworkers - Example: college graduates a ect the output of high school graduates, and vice versa (working together) - Workers contribution to output includes e ect on coworkers output. In this case college graduates employees contribute more to aboard productivity than high school graduates. - It pays to hire better quali ed L (labor) when output is interdependent. - Interior solution: employ a mix of “low-skilled” and “high skilled” labor. 3. Workers productivity is independent of coworkers but depends on capital - Example: production of a shirt involves both machines and labor, capital input can be changed (not xed) - New machines more than doubles the capital cost (7.50 for old machines and 16.50 for new machines), but does not double the total cost, may be cost e ective to use with professionals. fi ff ff ff ff fi 4 ff - There are di erent ways in which the labor types H and C may enter the production function: - Solution: - Corner Solution: - Hire LH if productivity contribution is larger than the cost associated with them (cost ratio). - Hire LC if productivity contribution is lower than the cost associated with them (cost ratio). - How many workers to hire? - To maximize pro t any resource can be used, including employees, up to the point where the marginal bene ts just equal the marginal costs if there is an interior solution (MB = MC). 1.4 Decision Making with Imperfect Information - Information required for analysis is often not immediately available or is expensive to obtain. There are di erent ways of making the decision: - Make a decision independent of the analysis: make the decision based on good felling and best guesses using instinct, experience or standard practice. - Estimate the relevant information for a formal analysis: simulation of di erent scenarios, which are more likely to lead to a better decision making than a simple guess but include costs of estimation. - Experiment and compare outputs of di erent groups. Example: hire di erent worker types and see how they perform. This is potentially costly, but depending on the situation, the returns might be high. - What are we trying to learn? (starting point) - Will the answer have an e ect on pro t (large/small)? If it’s small and costly time consuming, it should be dropped, but if it can potentially have large e ect, an analysis should be made. - What kind of data do we need to answer that question? (Important to specify) - What is the cost of collecting the data? - How reliable will the results based on the collected data be? If data is not very reliable, probably the whole data collection isn’t worth it. - Bene ts from economic theory: - Formal thinking guides the experience - Formal analysis helps tp improve decision making by clarifying the important (and unimportant) issues. - Formal analysis considers the whole picture (best case). ff ff ff ff ff fi ff fi fi fi 5 Topic 2. Recruitment Example: Recruitment and internal labor market - The 8 hierarchical levels of management ranks in ACME: - Level 1 acts as a “port of entry” into the company (hire from outside the rm) - Levels 2-8 are mainly promoted from within employees (internal labor market) - Data suggest evidence of sorting: - 21% of the employees leave within 2 years - 35% leave within 4 years - About 40% stay more than 10 years - Job O ers and Attracted Applicants: rms have to see or estimate if the applicant is a good t for the o ered job. O ering a high level of pay or bene t causes the problem of adverse selection due to asymmetric information. - Adverse selection: due to asymmetric information, the rm might not be able to di erentiate between applicants of low or high quality. This is due to fact that rms attract wrong kind of workers as well as desired ones. The rm could be able to get some information about the applicants but not as much as the applicants can have their-self. In order to avoid adverse selection, the rm could rely on credentials or learn about the productivity of workers productivity by screening (probation). - Credentials (signaling): - The most important credentials are the type of experience, type of training and quality of school. - A credential is informative if there is a positive correlation between performance on the job (productivity expectation) and the ability to obtain the credential. Credentials may provide two kinds of information about the holder: - Speci c ability (knowledge or skills) needed in the job. - General ability to acquire new knowledge for the job (languages, programming) - Property of credentials: there are di erent costs for obtaining the credentials (administrative fees and course and exams prices). An e ective credential is a credential that only most quali ed people posses and most low quali ed don’t (credentials that not everybody has). - Return of investment in credential: if the di erence in wages between people with di erent types of credentials (university/high school) is very small, there are not many incentives to invest in a university credential. Only the best quali ed people (for which is easy to obtain a university certi cation) would obtain the credential (lowest cost of obtaining it). - Signaling is one way to resolve adverse selection problems. Independently of the credentials needed to performance a job, the applicant can signal (make the rm notice) his abilities and credentials. - Learning workers productivity (screening). fi ff fi ff fi fi ff fi fi fi fi ff fi fi fi ff ff fi fi ff ff fi 6 2.1 Screening Job Applicants - The problem of homogeneity in applicants: after self-selection of applicants and sorting by information from their resumes, there are further screening methods for rms: - Job applicant tests: works better for xed and measurable tasks (can be manipulates). Applicants usually have to do a small case of study in order to show their problem solving skills and productivity. If this case of study is made online, the applicants could ask for help or even work in groups). - Phycological pro ling: survey with questions about the phycological characteristics of the applicants such as responsibility. Phycological pro ling is an inexact science and can be manipulated by the applicants by answering the questions with desired answers. - Personal interviews of job applicant: from simple interviews (meeting for asking general questions and get a rst impression) to more elaborate interviews (various rounds, in groups, performance of tasks or elaborate of tests). This screening method is the most time and cost intensive. Example: Pro tability of screening at di erent banks - Assumptions: - There are 5 hypothetical types of job applicants (A-E) with di erent levels of productivity. - There are 2 banks (Investment Bank and Commercial Bank), and both banks pay about 100000€ to all of the remaining applicants. - The screening tests costs 2000€ per person but gives a de nitive information about the type of applicant. Applicants Type A Applicants % Productivity Investment Bank (in thousand €) B C D E 10 % 20 % 40 % 20 % 10 % -250 € 0€ 125 € 200 € 450 € 95 € 100 € 110 € 120 € 125 € Productivity Commercial Bank (in thousand €) - This table shows how many applicants apply to both banks. The productivity indicator in the Investment Bank shows that there’s a large variation of productivity across the di erent types of applicants than between types of applicant in the Commercial Bank. - The worst scenario of the Investment Bank would be to hire applicants of type A or B, which generates negative or non productivity respectively. Knowing the wage is 100000€, the bank won’t hire an applicant whose productivity output is less than the wage of the worker (this leaves 70% of the applicants remain). - In the case of the Commercial Bank, the worst scenario would be hiring applicants of type A, which generates 95000€. Knowing the wage for the new workers is 100000€, the bank won't hire an applicant whose productivity output is less than the own wage of the worker (only 90% of the applicants remains). Bank Screening Cost Average productivity Pro t per worker No screen 110 € 10 € Investment Bank (in thousand €) 2857 € 193 € 90,10 € Commercial Bank (in thousand €) 2222 € 112 € 120 € Both Banks (in thousand €) - For the Investment Bank, as the screening costs for each applicant is 2000€, but only accept 70%, would pay on average 2857€ (2000€/0.7). For each applicant, the screening cost would be 2857€ taking into account that the bank is not going to hire a non pro table 30% of the applicants but would have to screen all of them. - For the Commercial Bank, as the screening costs are 2000€ per applicant, but accepts 90% of them, would pay on average 2222€ (2000€/0.9). For each applicant, the screening cost would be 2222€, taking into account that the bank would not hire 10% of the total applicants but need to screen all of them. Screening cost per person = Screening cost/ Applicants % Pro t per person = Productivity - Wage - Screening Average productivity = Total productivity of applicants / Applicants % ff ff fi fi ff fi fi fi fi fi fi fi fi 7 - Reasons for di erences in pro tability: - E ective test: screening is more pro table when tests are e ective. - High downside risk: screening more pro table the greater downside risk from hiring wrong person. - Pro tability of Screening: - Screening may not always be pro table to the employer because competitive labor markets would raise the wage of high-ability workers to their productivity. Companies might adapt the wages in relation to the productivity of the workers because if not, the high-productive worker would go to other companies. - Job applicants risk screening if the rm compensates for trouble and risk when getting employed. - When the rm cannot bene t from the screen, job applicants will have to pay much or all the screening: - Pre-job market screening (signaling: education or professional certi cations) - Lower pay during the screening period (probation) * It is likely that both, the employer and the employee will share bene ts and costs of screening. Probation - Probation: hire workers for a xed period, and only move on to create a long-term contract if the performance of the worker is adequate. The main problem of probation arise when there are termination costs (cost of ring), which can be substantial for the company. Probation is a general trend because of increasing turnover costs due to a greater employment regulation and litigation over wrongful termination. This problem of termination costs can be solved by hiring on temporary basis through temporary employment agency or (repeated) xed term contracts. - Implications of probationary screening: - Up-or-out promotion system: either applicants go up in the company and get a long-term contract, or they are terminated and get out of the company. - The rm is likely to promote those who survived the screening. Those who had a better performance in the probation time are sometimes compensated by having an increase in their wages with the new contract. Large raise in compensation will usually accompany the promotion. Formally Screening - Assumptions: - E and D are 2 di erent types of applicants with productivity Q (type E are more productive than type D). - The rm pays wage w. This wage is the same for both applicants, but generates pro t from type E - applicants, because the wage is smaller than their productivity output; and generates loss from type D applicants because the wage is greater than the productivity output. QE > QD Productivity of E is greater than productivity of D QE > w > QD Productivity of E (pro t) productivity of D (loss) P(E) = p Probability that a random job applicant is type E P(D) = (1-p) Probability that a random job applicant is type D q Probability that screening reaches the correct result (type E) 1-q Probability that mistakenly type D applicants are hired -p (1 - q) (QE - w) Loss from mistakenly hire types D (reject E) ((1 - p) q (QD - w)) Gain from appropriately hire type E (reject D) - The test is more e ective when it's more accurate (s) or discriminating (>p). The more e ective the screening is, the more likely to get the correct result (hire type E), better for the pro t. If screening costs are high, pro t decreases, and if the probability of hiring type E is very large, there’s lower bene t from screening. fi ff fi fi fi fi fi fi ff fi fi fi fi fi fi fi fi ff ff ff fi fi fi fi fi ff 8 2.2 Signaling - Signaling involves screening in order to overcome adverse selection. Applicants know their type (productivity), and rms structure job o ers in a way to attract the best applicants (to get a better match). Types reveal their quality, type, by accepting or refusing the o er. Example: Recruiting for an Investment Bank - Assumptions: - Bank screened out types D and E by interviews, but would like to hire high-pro table types E. - Applicants know what type of employee they will be. - Outside alternatives: what they could earn in another company wD < wE. - Two periods: 1 performance evaluation for both types / 2 promoted workers stay in the rm. - The rm o ers an up-or-out promotion: - In the rst period, both types “post a bond” (accept a wage w1), that is lower than what both could earn in the labor market. Applicants accept this low wage because they know that if they survive screening, they would receive a higher wage that the labor market w1 < wD < wE. - In the second period, if screening shows: - High productivity: employees get a reward such as hire and promotion with a higher wage than the labor market w2 > wE. - Low productivity: employees are red and receive market wage wD The rm o ers an up-or-out promotion: - In the rst period, both types (A and E) “post a bond” (accept wage w1), that is lower than what both could earn in the labor market. They accept this low wage because they know that if they are able to survive screening, they would receive a higher wage than the one in the labor market w1 < wD < wE. - In the second period, if screening shows: - High productivity: employees get reward such as hire and promotion with a higher wage than the labor market w2 > wE. - Low productivity: employees are red and receive market wage wD. - Values of application varies from type E to type D. - High quality types signal their type by incurring cost. The signal is willingness to accept a wage that is lower than the market wage w1 < wE during the promotion period. The fact that someone invests in this kind of opportunity proves that the applicant knows is able to get to the 2nd period because has high probability to be rewarded by promotion and higher pay. fi fi ff fi fi ff ff ff fi fi fi fi fi 9 - There is a need of rigorous performance evaluation before awarding promotion. Low success rate to get a promotion for types D reduces the expected wage of a job compared to types E. - For achieving an e ective signal, there must be incentives: - Low-ability types must be discouraged by having a low probability of getting a promotion. - High-ability types must be motivated to apply by having high probability to get a promotion. - Costs and Bene ts: - The employee will pay for most of the cost of signaling, and enjoy most or all of the bene t of the signal: - Risk that they won’t get promoted (have a loss) - Reward after promotion, earning more than in the labor market (bene t). - The rm may also pay for part of the probation, and get some bene ts: - Costs from paying more wage than productivity to those who are promoted. - Bene ts from paying wages below productivity during probation period. - Further examples of signaling: - Warranty (companies believe that the product is quali ed enough so that you can return it if it is not). - Venture (risky) capitalists (investment of a lot of money because believing on its pro tability) - Pay for performance (employees are paid based on their productivity) - Education: high potential workers signal talents by investing in education. Screens are increasingly di cult to pass and students enroll for next level of schooling if screens seem achievable. Firms recognize signals, paying more to higher educated workers. General Property of Adverse Selection Models: - Lower quality types try to associate themselves with higher quality types - Higher quality types try to distinguish themselves from the lower quality types Separating vs. Pooling Equilibria: - Assumptions: - There are 2 types of junior accountants (quicks and slows) that want to invest in some training. If they complete the training, they become Certi ed Public Accountants. - Quicks are more productive than slows, which means that nd it easier to pass the CPA exam. - Labor market pays accountants a wage that equals their expected productivity. - Model: - Average productivity without signaling: - Conditions in order for signaling to work: Incentive for quicks to obtain the credential: the productivity of the quick minus the cost of obtaining the credential for quicks should be higher than the productivity of slows. Incentive for slows to not obtain the credential: the productivity of slows should be higher than productivity of quicks minus cost of obtaining credential for slows. The gain from signaling must be higher than the cost for high ability types, but not so high that low ability types are also motivated to signal. The gain of signaling must be higher than not signaling for all quicks. Since the expected productivity is higher than the productivity of the slows Q’ > Qs, the condition is stronger for quicks. The larger the fraction of quicks, the closer the expected productivity to the quicks (this would not hold last condition. - Pooling equilibrium: if the previous conditions are not met, it's better for the workers with low quality because quicks don’t distinguish themselves from slows. There would be no signaling because neither slows neither quicks has incentive to obtain the credential. - Separating equilibrium: if confqtiosn are met, it’s better for the employers/company because quicks are able to distinguish themselves from the slows. In this case quicks signal (invest in credential and slows don’t. - Which rm is more likely to use signaling? Signaling is useful when di erences in talent among potential employees matter a lot to productivity. Signaling is more important in jobs where skills are most important and there is not much information available about the job applicants. ffi fi fi ff fi fi fi fi fi ff fi fi fi fi 10 Topic 3. Investing in skills 3.1 Matching - Every lm is di erent in its business, organization, and corporate culture. Employees with similar abilities will not t equally well at the same employer because of di erences in what it is important for them. There is like a marriage market in the labor market where companies and employees get married (match). Hedonic Wage Function: Firms have di erent isopro t curves and di erent workers have di erent indi erence curves. Labor market marries workers who dislike risk with rms that nd it easy to provide safe environment, and workers who don’t mind risk very much with rms that nd it di cult to provide a safe environment. The observed relationship between wages and job characteristics is called a hedonic wage function. - If matching is important for a speci c job: - Employees and rms need to seek out good matches with each other. There is a search motor in the labor market that companies and employees need to learn from each other whether they are good t or not. - Companies don’t only need to seek workers regarding their abilities (signaling or experience), must be sorted in relation to their attributes (portfolio of skills, personality or locational preferences). - Matching may imply a high turnover in the early career, where both workers and employers (companies) learn what is important for them. - Outside hires (external labor market) would have greater/more variance in employment outcomes (attributes and turnover rates) than internal candidates (internal labor market). - Human Capital: - Human capital is de ned by OECD (Organization for Economic Cooperation Development) as knowledge, skills, competencies and attributes embodied in individuals that facilitate the creation of personal, social and economic well-being. - Can be acquired in many ways such as education or other pre-labor market training and on the job training. - Human Capital Theory: theory regarding why do people invest in education and training. Analyses investments in education and training, which are modeled just like any kind of investment. - Decision rule: investments are only made if: - The present value of the bene ts generated by the investment - Exceeds the present value of the investment costs. 3.2 Investment in Education (or other pre-labor market training) - Net Present Value of investment in education: 0: present year (year of decision whether to drop out or nish college) T: T future years or las year of career. r: interest rate per year Ht: earnings in future years if student drops out of college Kt: earnings in future years if student continues college C0: direct cost of education (tuition, textbooks, supplies) F0: indirect cost of education (opportunity cost) - Decision: - If NPV > 0, it would be better to nish college - If NPV < 0, it would be better to drop out of college and invest the potential fi fi fi ff fi fi fi fi ff fi fi fi fi fi ff ff fi fi ff ff ffi 11 - Implication: the lower the costs are, the higher the Net Present Value. - For early years of schooling, returns to schooling. Exceed the costs. - Foregone earnings during early years of schooling are very low. It pays for almost everyone to invest in (general and at least some speci c) education. - The stopping point is the point where NPV switches from positive to negative. - Factors that a ect the Net Present Value: - Costs: higher standard fees lower the number of students as ell as weaker work labor market opportunities (more people obtain an MBA-Master of Business Administration during a regression). If a person already has a high paying job, has a lower probability of inviting in education. - Interest rates: don’t change much in the long term. Borrowing rate of parents is low. - Career length: older people with shorter remaining career length invest less in schooling. - Specialization of human capital: higher increase of income, comparative advantage and gains from trade due to acquire specialized skills (not many people has expert skills). - E ectiveness of learning (ability, school quality): relates to the ability and the quality of learning. The easiest learning is for a person, the less time he needs to nish the degree (less money). 3.3 Investments in On-the-Job-Training - On-the-Job-Training: training during working. - Direct costs: books and any resources needed, and compensation for trainers. - Indirect costs: on the work training takes workers’s time and attention away from regular duties (lowering productivity). Putting a worker who is not fully trained into a job would have less productivity compared to putting a fully trained worker on the job. - If a person expects to stay at a current rm, should focus on the skills the employer ( rm) values most. - If person expects to leave, best strategy would be to invest in the skills the labor market values most. fi fi fi fi ff ff 12 labor market for the skills, many persons demand those skills, which can be acquired outside of the workplace and the person don’t need to be in a job for it. Can be used anywhere, which means that the company has no incentive to support the worker to acquire this training. - FHC (Firm-speci c Human Capital): means that training has no value outside the rm. Thin labor market, not many companies demand this certain skills (speci c), which are acquired in the workplace such as speci c production methods or intangible rm speci c knowledge about corporate culture. Can be used practically only in the company of the worker, which means that the company has an incentive to support the worker on acquiring this kind of training. - A special case is Intellectual Property, which has both general and rm-speci c investment. - Who should pay for training? - Education is GHC (General Human Capital): a worker should pay for investment in education. - Exceptions ( rms might pay some of the employees’ education costs): - Implicit quitting cost (if the person quits the job just after graduation or something similar, he would have to pay back the tuition costs to the rm). - Strong match between worker and rm (expectation of the employee to stay for many years). - O er speci c training in order to generate self-selection in recruiting (recruit people interested in acquiring the type os skills the company o ers. - Cost advantages of rm compared to employee (the rm can save some taxes if paying education of employees and can increase the productivity of the workers). - On-the-Job-Training: mix of general and rm-speci c human capital: - General Human Capital (GHC): worker should pay 100% of the investment and receive 100% of the bene ts from the investment. - Firm-speci c Human Capital (FHC): the payment is likely to be split up between rm and employee. - Holdup Problem: if one of the parties makes an investment and expects to earn bene ts later, but a second party is tempted to renegotiate after the investment is made. If the risk is foreseeable (predictable), the investment may not be made, for fear of loosing some or all of the returns if forced to renegotiate later. - For example worker accepting a lower wage while investing on training because after will receive a higher wage, there is the risk that the rm wants to renegotiate the wage after the training is done. - Solutions: rely on trustworthiness of parties (reputation) or split cost and return of investment. Investing in Firm-speci c Human Capital (FHC) The usual net productivity of the worker is 10000€. During training, the workers productivity gets lower to 5000€ because spends half time for training, so the worker and the rm should split the costs of this lower productivity, reduce wage. After the training, the productivity of the worker in the speci c company would increase till 15200€ and would earn 14400€, but as they shared costs before, the rm and the worker should split the return and the worker should get a higher wage. fi fi fi fi fi fi fi fi fi fi fi fi fi ff fi fi fi fi fi fi fi fi fi fi fi fi fi 13 ff fi - General Human Capital (GHC) vs. Firm-speci c Human Capital (FHM) - GHC (General Human Capital): means that training is valued equally inside and outside the rm. Thick Example: Control Data Institute, WIPRO Technologies - New hires get 3 months of training in software design (largely General Human Capital ) if they provide an interest free deposit of 1400€. The deposit is paid back if the workers stay with the rm for at least 12 months. - In this case, if the holdup problem of providing general human capital can be solved by a joint investment in rm speci c human capital, a relationship between employer and employee is generated. The connection between employer and employee becomes long term, both are interested in lower turnover, and internal labor markets become relevant. - Assumptions, investment in a mix of skills: - On the job training in skills: programming in Java (J) and knowledge of Tax laws (T). - The cost of the investment is 1/2 (J2 + T2). This implies increasing marginal cost of investment in skills J or T, so it is cheaper to invest in a mix of the two skills than to invest a lot in one of these skills. - The weight that rms gives to skill J is λ, while to skill T is (1 - λ). - The potential wage of the rm: w = λJ + (1 - λ) T - 2 period model: - Period 1 is for investment on skills (decide for a certain type of training). - Period 2: with probability p worker stays at the company with weights on skills λ. With a probability (1 - p) leaves the rm and works at other rm with di erent weights on skills ^λ. - Expected Wage in Period 2: maximize the expected wage with respect to J and T, taking p, λ and ^λ as given. - First order conditions: - The optimal investment is a weighted average of skill-values inside and outside the rm. - If the workers nd likely to leave within the next period, the change in earnings from switching employers decreases. Decreases with the probability that the worker will leave the rm next period. - An increase in market thickness (more companies demand certain skills) leads the worker to invest in a way that is more consistent with the current rm’s relative valuation of skills ( rm-speci city of human capital is endogenous with respect to market thickness). - With more than one period remaining (after training), employees will invest less in rm-speci c human capital because the probability that the worker will move to another employer (company) is higher. Example: Investment in a mix of skills 1. Continuation is certain: p=1; λ=1/3; w (current wage). J* = pλ + (1 - p)λ^ = 1·1/3 + (1 - 1)·λ^ = 1/3 T* = p (1- λ) + (1 - p)(1 - λ^) = 1·(1 - 1/3) + (1 - 1)·(1 - λ^) = 1 - 1/3 = 2/3 w = p [λJ* + (1 - λ)T*] = 1·[1/3·1/3 + (1 - 1/3)·2/3] = 1/9 + 2/3·2/3 = 1/9 + 4/9 = 5/9 2. With those investments you would have to switch to a rm with λ’=1/6; p=0 (separation is certain) w’= (1 - p)·[λ’J* + (1 - λ’)T*] = (1 - 0)·[1·1/3 + (1 - 1/6)·2/3] = 1/3 + 5/6 · 2/3 = 1/18 + 10/18 = 11/18 3. Suppose you made your investments believing you would work for a rm with skill demand: λ’=1/6; p=1 J* = pλ + (1 - p)λ^ = 1·1/6 + (1 - 1)·λ^ = 1/6 T* = p (1- λ) + (1 - p)(1 - λ^) = 1·(1 - 1/6) + (1 - 1)·(1 - λ^) = 1 - 1/6 = 5/6 w’’ = (1 - p)·[λ’J* + (1 - λ’)T*] = 1·[1·1/3 + (1 - 1/6)·2/3] = 1/3 + 5/6 · 2/3 = 1/18 + 10/18 = 11/18 fi fi fi fi fi fi fi fi fi fi ff fi fi fi fi fi fi fi 14 Topic 4. Managing Turnover - Turnover: workforce ows in and out of the rm. The rate at which employees leave a company and are replaced. Job-Turnover rate in West (ABL) and East (NBL) Germany Before 1990, East Germany (NBL) belonged to the former soviet union and was under a completely di erent political and economic regime. Due to the restructuring and integration of the federal German states in the early 90s, was a huge change for companies sector because new companies appeared and some old companies had to shut down. Due to this fact, the Turnover in East Germany was very high and started decreasing due to the appearance of new companies and “opportunities” for the workers to change their jobs. Job-turnover rate vs. Labor-turnover rate - Job-turnover refers to the total share that the rm has. - Labor-turnover re ects how many employees actually change (were structured) in the company. - Example: A company employs 100 individuals in t=0 and 110 in t=1 - Job-turnover = 10% - Labor turnover = 10% A company employs 100 individuals in t=0 and 110 in t=1, additionally 10 individuals left the establishment and 10 were hired to replace them. - Job-turnover = 10% - Labor-turnover = 30% 4.1 Is Employee Turnover good or bad? - Turnover due to di erent circumstances: - Regular workface ows: - Recruitment: when new persons get into the company. - Promotion: when promoting persons between levels the company would have turnover workforce ows. - Other regular workforce ows are dismissals, resignation, terminated contracts or death. - Irregular workforce ows due to downsizing: - Layo s: when companies close down and dismiss around 50% of their workers. - Buyouts: when there is a take over by other company and people are massively laid o . ff fl ff fi fi fl fl fl fl ff fl ff 15 - Factors a ecting rm’s optimal turnover in normal conditions: - Importance of sorting: select the correct applicants for the job. - Technical change: changes the way that employees do their jobs and not all might be skilled prepared (replaces jobs but also creates new opportunities). - Organizational change: organizations reactions to dynamic environment and adaptation to be able to respond to new demands. Reduction of employees due to nancial losses or for simpler communication. - Hierarchical structure: changes on the hierarchical structure for instance, to make communication easier. - Speci c human capital: depending on the needs of the company, that might also change, there would be more or less turnover ows due to this changes. - Tools to reduce turnover (retention strategies to prevent employees from leaving): - Increase compensation: increase the wages of the workers (might be simple but expensive). - Treat key employees as partners: promoting workers to levels where would have more responsibilities or in uence. In case one of them leave after having promoted him would be very damaging for the company. - Meet the speci c taste of the worker: what does the worker want and which is important for him like exible working hours or new opportunities such as working on a di erent eld on the company. - Paying attention to implicit contracting issues: if the implicit promises in the contract does not t the employee, this one might be disappointed, so should be a good idea to reduce disappointment and invest in positive reputation. - Reducing cost of turnover: - Noncompete agreements: if a person wants to leave the company there might be in agreement in order to that person don’t steal information or customers. - Worker collaboration: teamwork between employees. Cross-train employees or job rotation: task or place changing between the employees. Job standardization: the job is structured in a way that other person can do it without being completely lost. Knowledge management: knowledge base where all that the company and the employees know is. There is an agenda where the company can nd who to count on for which kind of tasks. - When is turnover good? - Professional service rms: turnover makes sense and is an essential part of their business model. Use of the up-or-out promotion as a strategy to nd out which employees really matches the company. - Motivational tool in volatile industries: the quicker the industry changes, the quicker the company needs to respond to the changes. It makes sense to have some exibility in terms of turnover in order to increase innovation and “free” training (might train some employees and get some new already trained from other rms). - Reputation building: attraction of new high quality employees and dismiss not suitable employees. 4.2 Bidding for Employees - Wages (pay) determination in perfect competitive markets: - Requirements for a perfect competitive market: full information, complete contracts, in nitely many buyers and sellers, and no transaction cost. - Productivity Q must be higher than the pay or wage w, which is higher than reservation wage R (minimum wage that the worker requires in order to participate in the labor market). Q≥w≥R - Winner’s curse in Raiding (steal an employee from another company): - If an individual has speci c skills that are really found there is no other option and hiring from another company might be even necessary or the only way to hire that person. But, if the individual has skills that are commonly found through the market in other employees, it would be better to hire from a pool of applicants and not from another company due information asymmetry. - Disadvantage of hiring from another rm: information asymmetry, another rm would have more information about the workers productivity than does an outsider company. - When to raid a rm for an employee? Criteria: - The worker’s value is greater to the raider than to the current rm of the worker. - The worker’s current rm does not overvalue and therefor, overpay the worker. Productivity has to be higher than the received wage (Q > w). fi fl fi fi fi fi fi ff fl fi fi fi fi fi fl fi fi fi fi fi ff fi fl 16 - Raiding could be pro table if: - There are recently completed schooling (degrees) such as university or acquired technical skills. - Is used in volatile industries, if environment, organizational or structural changes occur very often. - Is used in industries with rapid technical change: specially IT software. And hardware companies where not all the employees are skills-updated. Bene ts for the employee of searching and new job: - NPV: Net Present Value for the employee T: years the wage increase can be utilized r: interest rate wi: wage increase p: probability of nding a higher paying job c: search cost - O er matching policies: if a worker gets a better o er for a job than the one at his current rm, he can try to get a better position or increase the wage if the company decides to match the policies in order to retain the worker. - The rm can follow next strategies of policy matching: - With o er matching: if the rm wants to retain the worker, it would o er matching policies up to the point where the wage equals productivity (w = Q). - No o er matching: can discourage workers with strong preference for the current rm. - Selective counter o ers: train exibility depending on the person and the replacement cost. - Non monetary compensations: health insurance, daycare for children, gym, rets areas, status, exibility, location, job security. This compensations are hard to evaluate and bene ts to search for the employee but irrelevant if he is just “for the money”. fl fi fi fi ff ff fl fi fi ff fi fi ff ff fi ff 17 4.3 Layo s and Buyouts - Who to target for layo s: rst thought might be to get rid o of most expensive employees, but most expensive are normally the most productive too. The indicator relevant for the decision is the performance to cost ratio. - Firm maximizes pro ts by laying o from both ends of age distributions the rst (young and old workers, people that are about to leave the company anyway). - Medium age workers are most productive. t: time Kt: productivity at the rm Wt: wage at the rm At: alternative outside the rm Pt: amount of pro t to the rm - Cost of layo s: - For younger workers the LIFO (Last in First out) principle might be applied because this workers are the ones the company invested the less in terms of training. - For older workers is more controversial due to anti-discrimination laws and implicit contracting. If people though they were going to spend their whole working life on the company and the rm terminate them it could lead to a reputation problem or crisis if they telco other people and new employees. - Litigation due to wrongful termination. - Buyouts: payment make for a worker to leave the company. - The buyout compensation is usually tied to: - A working not suiting the company - For not public scrutiny (to not make a scandal about the hire) - O er for training a new worker (to train the person is going to replace the old worker) - By o ering buyouts, a company can face the problem of adverse selection (the one that leaves the company can often be the most productive workers). - A rm o ers a buyout if the present value (cost) of the worker staying (wage) is greater than his productivity. - Worker accepts buyout if it’s greater than the value of staying minus the value of the alternative. Would accept if the buyout amount is greater than the one he would receive if he stays in the company (di erence in wages). - A deal can be made if the di erence between wages and productivity from the company side view is more or equal to what the worker has as a minimum buyout in order to accept. When the value of productivity is larger than the alternative wage in other rm. Firm o ers buyout if: Worker accepts buyout if: PV(K): Present Value of productivity PV(W):Present Value if worker stays (wage) Buyout deal making if: PV(A): Present Value of alternative (wage) - Implementation of layo s and buyouts: - Has to be a quick, surprising and extensive process. - Increase the possibility of layo s, if workers don’t want to accept the buyout just threaten with a layo . - Window-plans: o ering a buyout for a short amount of time. - Early retirement with full pension or retirement bridge: seniority credit to make up for money for pension if they had kept working. - O er job placement service: nd new job for the worker leaving the company for not being unemployed. ff ff fi ff fi ff fi fi ff fi fi fi ff ff fi fi ff fi fi ff ff ff ff ff ff fi ff ff 18 Organizational and Job Design Topic 5. Decision Making 5.1 Organization of an Economy - Debate of the 20th Century: How to organize an economy? - Centralized way (largely run by the government), with government intervention. - Decentralized way (less run by the government), with less government intervention. - Decentralized market-oriented economies are much more e cient because the stakeholders can make decisions and optimize the market. Decentralized markets are better creating growth, jobs, and overall prosperity; and are more creative and adaptable (react quicker to changes in the market). The Market as self-organizing system (Adam Smith), the invisible hand leads individual sel sh decisions. - Bene ts of local planning: - Markets as Information Systems (Friedrich von Hayek): markets are a form of collective intelligence because every person takes decisions and this decisions maximize each ones welfare. - Information is local and idiosyncratic: people know what is the best to do and prices implicitly take into account this dispersed knowledge. - In decision making: prices are the most e cient way to carry information of “particular circumstances of time and place”, which is needed to coordinate resource allocation across industries and even countries. - Whenever a person has goods produced, he depends on how much of the goods exist (supply) and how much demand of the goods it is on the market. - Prices are an economic information system (because depends on supply and demand). - Markets as Incentive Systems: asset owners have a very high incentive to survive the decisions they make (receive a high pro t from their own decisions). A central planner/bureaucrat, less attach to the assets, has less incentives to use this assets e ciently. - Market gives strong incentives to use the assets e ciently (against state intervention / centralization). - Market economy incentives imply that resources tend to be allocated to the “right” market participants. This resources are allocated with the individuals who got the information or skills that are most valuable. - Markets and Innovation: since markets are information and incentive systems, asset owners are motivated to respond to problems and opportunities quickly and e ectively, and have incentives to invest and create new pro table products or services. - Decentralization allows economy to use creativity and ideas of all individuals throughout the society. - Centralization use only ideas of those who have access to central planner, the rest are ignored. - Markets are a great source of innovation and adaption, can react more exible to local circumstances. - Bene ts of central planning: - Examples for market failures: - Economies of scale, which lead to natural monopoly: this economics of scale usually arise if there are large xed costs, and the cost per unit decline as the rm gets larger. Governments usually regulate monopolies in order to reduce ine ciency. - Public good problem (lake, parks, beach): consumption of this goods cannot be charged or prevented and, if the value of the good is high, government might provide it (take care for it and pay maintenance). - Externalities: - Negative externalities (pollution by companies): might be reduced by with the imposition of taxes. - Positive externalities (standards): government might impose subsidies if the standards are valuable but, by coordinating the standards too much competition is reduced. - However, competition for nding best standard might be bene cial for creating better technology, so it is not clear if the central decision making is necessarily good. - Market Metaphor of Organizational Design: markets and rms organization are in need of same important functions (similar organization). Tasks to develop a model of organizational design: - Use central and local knowledge (strong management vs. empowering workers) to increase e ciency. - Coordinate decisions when necessary, like brand name or logo. - Provide incentives to make coordinated decisions, innovate and adapt * Market prices are better measures of value than performance measures (market metaphor is imperfect). ffi fl ffi fi fi ff fi ffi ffi fi ffi ffi fi fi fi fi fi fi 19 5.2 Bene ts of Centralization in a Firm - Common assets: have a public good character and economies of scale (headquarters) for the whole organization, centralized planning and decisions, common accounting system with headquarters that share overhead costs. - Strong brand name with centralized decision about product lines. - Strong reputation with centralized human resource policies to increase it. - Central knowledge: combined experience of the entire rm regarding strategies. Centralized decisions due to patterns and trends (overall strategy) and possibility of improved transfer of knowledge across units. - Coordination problems: - Assembly lines: can only move as fast as the slowest link so the speed is determined by the task that takes longer, and units need to be synchronized (military). - Externalities coordination problems, Research and Development investments (asymmetric information). - Strategic decisions often need to be centralized because involve the whole company and products need a level of compatibility. If products don’t need to work with each other, strategic decisions can be decentralized. 5.3 Bene ts of Decentralization in a Firm: - Communication cost of speci c (local) knowledge: speci c and technical knowledge that is costly to transmit or understand makes decentralization favorable. It’s easier and less costly to leave the decision making to the people that know how to manage the speci c information for making the decision. - Perishable (temporal) information: quick decisions are needed due to the temporary value of the information. Complex information: interdependence between di erent pieces of information (cannot be put together). Information that requires technical skills: need to transmit the skills to understand the information. Unforeseeable or idiosyncratic information: information in a random or dynamic (changing) environment. Subjective or experimental information: information di cult to quantify or describe. - Save management time: less important and less coordination-dependent decisions are taken at the lower levels of the rm (down the company’s hierarchy) so that management can concentrate on its core issues. - Develop management skills: train the analysis and decision-making skills of the employees that could be promoted into higher positions in the company later on. - Lead to job enrichment: employees with higher intrinsic motivation on the job due to more interesting and challenging tasks as well as more responsibilities and participation on the decision-making. 5.4 Authority and Responsibility - Job design determine the degree of authority and responsibility associated with a given job, which depends the job holder. In a team, one of the members must have the authority to make the decision. - Decision making as a multistage process: - Stages: 1. Initiatives: brainstorming of possible options. 2. Rati cation: choose or select options and strategies. 3. Implementation: implement the chosen strategy, put it into practice. 4. Monitoring: control if the implemented strategy really conforms the chosen strategy. - Responsibilities of stages: stages have di erent levels of responsibility and can be given to di erent persons. - Decision Management: Initiatives and Implementation are part of the decision management generally more decentralized. The speci c knowledge needed for this stages is available at lower levels in the rm. - Decision Control: Rati cation and Monitoring are part of the decision control typically centralized. - Information and decisions up/down the company operate by hierarchy due to the need of coordination with other rm strategies and incentives of employees to favor their individual goals. - Flat vs. Hierarchical Structures: - Flat structure: emphasize decision management, everybody take part on the decision process. - Steep (hierarchical) structure: emphasizes decision control, there is a department that takes over the monitor process (not everybody participates on the decision-making process). - Job design with at or steep authority structures depends on the costs of the respective errors. fi ff fi fi ffi ff ff fi fi fi fi fl fi fi fi fi fi 20 - Types of errors: - False negative error: when a production line is pro table but the company decides not to produce. - False positive error: when a production line isn’t pro table but company decides to produce anyway. Tradeo between false positive and false negative errors: Product Policy Aggressive policy Conservative policy Likelihood of false positive error High Low Likelihood of false negative error Low High - To reduce costs of decision errors: - Hierarchical structure: - Reduces false positive errors but increases false negative errors. - More approvals are needed in order to go ahead (emphasizes on decision control). - Takes more time because more people are involved, so fewer projects are evaluated. - Flat structure: - Reduces false negative errors but increases false positive errors. - Every project evaluated by only one person (emphasizes on decision management) and considers more projects than the hierarchical structure. - Flat structure with second opinion required: - Is less stringent than hierarchical structures but more than at structures. - Every project is evaluated by 2 persons (second opinion). - Approves more projects than hierarchical structure but considers less projects than at structure. * The structure that should be adopted by a rm depends on the risk structure Risk structures: - Mistake can be disastrous (oil drilling rm): hierarchical decision making is preferred in order to avoid costs of false positive errors. Is better to avoid something that can turn to be bad (false positive) than avoiding something that can turn to be good (false negative). - Little to lose (young startup): at structure preferred in order to take chances. - Symmetric payo s (most rms): tolerable levels of both kind of errors preferred. - Investing in better quality decision making improve the decision process: hire better evaluators (priced workers), give the di erent evaluators more time to consider each project and make more information available to evaluators (hire an outside consultant or buy data). fl fl fi fi fi fi fl fi ff ff ff 21 Topic 6. Organizational Design - Hierarchy: communication, supervision and decision making in a clear line. - Use the idea of a single decision maker (one person makes the nal decision). - Costs increase as the rm gets larger, mainly because of communication costs. - Information (relevant to make decisions) might get lost in the hierarchy (don’t arrive all levels of the company). - Types of Hierarchy: - Functional Hierarchy: - Groups of employees with similar tasks and skills. - Careers tend to happen within a certain function: a worker might promote to higher levels of its own background (sales, production or R&D). - Utilizes the bene ts from specialization: people working together own similar skills and have a similar knowledge, which makes the communication easier. - Narrow job design that provides better. Control of what happens on a certain function. - With increase of companies size it gets harder to coordinate decisions, which implies higher costs. With more people (and more levels/layers) it gets harder to transfer the information through all the levels. - This hierarchy disincentive for future innovations because there is no interaction, coordination or communication between functions. - Divisional Structure: - Divisions with functional structures inside the di erent divisions. As the company grows additional substructures are needed. Each division might use a functional structure on its substructures. - Specialization on a larger scale (country or product-line). Implemented for the need of di erentiation between product lines or bigger scales such as geographically division. - There are positive and negative externalities between divisions. - Loss of specialization and economics of scale. - Disincentives due to no interaction, coordination or communication between the di erent divisions. - Matrix Structure: - Structure that avoids miscommunication and combines the function and division view. - Balances the need for specialization and economies of scale - Multiply the career paths (workers can be promoted throughout the di erent levels of the company). - Communication at lower levels is increased. - There are two decision makers. - This structure is likely to increase o ce politics (try to get promotion) and bureaucracy. - Network Structure: - Team based structure governed by informal relationships. There is not a chain of command to follow so communication goes throughout he company without an imposed hierarchy. - Structure of quick decision making and better communication. Has similar bene ts as the matrix structure and utilizes the bene ts from specialization. Usually there is no pure implementation because there is no clear decision maker. This structure is an imperfect substitute of the market design. - Coordination problems: - Synchronization problem: there is no need for actual communication but the output of the workers must be synchronized (assembly lines or strong brands). There has to be a continuous rhythm of production for the company to work well. - Integration problem: in order to produce the nal product or service, the company needs to combine multiple pieces of speci c knowledge. It would be e cient to decide in a decentralized way but it is not possible due to externalities. If centralization is possible there is a high communication cost because there might be a lateral communication between the lower levels (matrix structure). One example is the design of a laptop. - Coordination mechanisms: central budgeting/planning, training/standard operating procedures, corporate culture, communication, general manager, job rotation, personality, networks, performance, evaluation and incentives. ff ff ff fi fi ff fi ffi ffi fi fi fi fi 22 Functional Hierarchy Divisional Structure Matrix Structure Network Structure Emphasize decision control Emphasize decision management - Factors span of control: routine (+) / complex (-) tasks, individuals skills (+), incentive plans/monitoring (+) and IT (+). - Skills, pay and structure: additional levels as incentive plans to keep high skilled workers, job title as compensation and general development in the structure of a rm. - Start-up: at network structure - Established rm: steep formal structure - Diseconomies of scale: larger rms and more employees, supervision, management, levels or divisions and distance between the CEO and the lower levels. - Diseconomies of scope: greater specialization and more divisions, managers, communication and more severe with di erent complex divisions. fi fi fi fl ff 23 Topic 7. Job Design 7.1 What is a Job? - Job: employee-employer match, worker provides labor (input) and the company gives a remuneration, normally money, in exchange for the services. - Job can be de ned as a set of tasks and is characterized by (dimensions): - Multitasking: number of tasks and task variety in the job. - Skills: required human capital for the job - Discretion: degree of decentralization/independence of a person’s performance and how closely is supervised. - Interdependence: importance of coordination (requirement) with other jobs or processes in the company. - Job design is the “art” of bundling tasks into jobs (how to distribute tasks across the di erent jobs in a company). - Goals of job design: - Improve worker’s e ciency in performing tasks (gains from specialization). - Create and make use of worker’s knowledge on the job (gains from innovation): workers coming up with new ideas that could help on the decision making process. - Improve and keep worker’s motivation. - Types of job: - Narrow job: normally is 1 task job (not multitasking) that requires low skill level, without discretion and no interdependence (centralized performance requiring coordination). - Enriched job: multitasking job that requires high skill level, discrete and interdependent. * Over time, jobs become more enriched, increases in one dimension tend to be accompanied by increases in others. 7.2 Multitasking, Multiskilling and Specialization - Multitasking: situation in which a worker is required to perform a number of di erent tasks. - Multiskilling: situation in which a worker has the ability to perform a number of tasks. - Makes worker more exible (valuable for small rms with few workers cause there’s no possible specialization). - Facilitates communication within teams or between departments (in need of coordination it is an advantage for workers to have some knowledge about the others job). - May assist making innovations on jobs because if a person is capable to perform many tasks, it’s more likely to come up with improvements or innovations. Specialized workers might be a bit reluctant to come up with innovations because those innovations might lead to the avoidance of their own jobs. - Specialization: situation in which a worker is required or specializes to perform only one or a few tasks. - Advantages of specialization: - Over time, workers can perfectly perform their narrow tasks (specialized human capital investments). - There is no loss of time and mental focus because there is no task switching. - Breaking the process down into narrow tasks, and giving workers specialized jobs, can result in enormous increases in productivity. Any move away from specialization should have a strong justi cation. Nail Factory Example (Adam Smith): - Making a single nail involved up to 18 di erent tasks and specialized factories may employee 10 di erent types of workers to perform this tasks. According to Smith, total daily output: 10 Multitasking independent workers: 200 nails/day 10 Specialized workers as a team: 40000 nails/day ff fi ff ff fi ff fl ffi fi 24 7.3 Finding the right approach - Multitasking (Advantages): - Lower transaction costs because the multitasker can respond quicker to the clients demands because the whole process is under the control of one person. - Complementary tasks: tasks A and B might be complements in production, meaning that any improvement in performing task A raises output in task B. A repair work involves 2 tasks, the diagnosis of the problem and the actual repair, which are complementary because the repair would be well done only if the diagnosis was too. - Interdependent or complementary tasks should be done by one multitasking person (modularization) cause specialization would create high coordination costs (monitoring costs, delays, low motivation or missed innovations) between the workers that are given the di erent tasks. - Tasks should be bundled in a way that gains from complementarities are maximized and transaction and other costs are minimized. - Classical Job Design (Taylorism): industrial engineers came out with best practice or way of performing a job and workers had to follow exactly this invented procedures. Centralized jobs, very narrow (few tasks), low skills required and low discretion (poorly motivating). - Modern Job Design (Continuous Improvement): part of Total Quality Management (TQM) - Constant evaluation and try to improve the process by involving the knowledge of all the workers. - Use ideas from the talents of the existing workforce as a source of knowledge, and oppose to use research, consultants or equipment (external experts), which would be very expensive. This use of ideas of the workers is a good way to motivate workers. - Improvements are not likely to be giant leaps (jumps), and therefore they are easier to implement (it is easier to optimize simple processes). - Intrinsic motivation of performing a task, and therefore job design, depend on the level of: - Skill variety (multiskilling), task identity (emotional relationship with job) and task signi cance (job importance). - Autonomy (interdependence) - Feedback (knowledge about the actual results of the work). - The Classical approach emphasizes pro ciency (being good in performing tasks): - Employees can be e cient but might have some motivation problems. - E ective for optimization of speci c tasks, control and uniformity. - Requires employees with lower skills. - The Modern approach emphasizes employee learning: - Employees have better motivation. - E ective for adaptation, change implementation and customization. - Require more engaged employees with problem-solving skills. - Finding the right approach (right mix) - Specialization combined with teams (team as multitasker) - Decentralization of some stages of decisions (lower levels) and centralization on others (higher levels). - Ex-ante optimization, standard operating procedures, automation and computerization; and continuous improvement, job enrichment, and deep/broad skills. fi ff fi fi ffi ff ff 25 Topic 8. Advanced Job Design 8.1 Teams: bene ts and costs - Bene ts of teams: - Complementary team members allow member specialization and, “ whole is greater than sum of its parts”. - Possible knowledge transfer among team members (team members become more productive). Knowledge transfer means that one person / worker gives part of his knowledge to another worker. The transfer can only occur if both workers or team members aren’t too specialized (di erent) and if other member’s human capital is relevant for task execution. - Knowledge transfer would not occur if there is too much overlap in human capital (both members have same tasks or knowledge) or if other worker’s human capital or knowledge is irrelevant for the other. Tasks A Knowledge transfer Tasks B - Costs of teams: - Slower decision making due to “bureaucracy” (long unproductive meetings) and di culty of consensus. The more democratic the decision is made, the more lengthy and slowly the decision making process might be. - Free riding problem arises. If an individual putting the e ort to increase or improve the team result (output) is paid by performance of the team, the individual workers might reduce their e ort. This is because the worker who does a greater e ort would earn the same amount as it teammates (would not earn full marginal bene t). Example: Fishing Industry (Teams) Activity Bene t Costs Remarks Select shing site 2 2 Multiple useful judgements but slow decision making. Fishing on small boat 3 0 Collective performance of tasks and low monitoring costs. Fishing on large boat 4 1 Independent tasks (free riding problem may arise). Selling the sh 0 4 Sales as independent task (di cult monitoring) 8.2 Teams: implementation and composition - Team size: - Small team size: - Advantages: easier coordination and monitoring, which reduces the free riding problem. The free riding problem decreases in small teams because: - Shirking (free riding) is more likely to be detected (due to easy monitoring). - Incentives to punish shirkers (free riders) is decreasing while the team size grows (the bigger the team, the less the punish). - Disadvantages: less opportunities for knowledge transfer and for economies of scope (the production of a good reduces the cost of producing other related good). - Large time size: in larger groups, instead of monitoring to reduce the free riding problem, there might be some norms stablished with the aim to reduce this free-riding problem. - Norms: informal policies/practices or set of beliefs held by the majority of a group. Common practices and beliefs can help establishing culture, ethic or etiquette in groups. - Norms and corporate identity may help to reduce free riding in large teams or rms and a ect worker’s behavior by creating empathy, loyalty and talent guilt. - Costs of establishing norms: - Creation cost of norms: introduction of the norm (working on Saturdays). - Maintaining cost: - Enforcing cost (penalization of those who don’t follow the norms). - Labor cost increase or additional labour cost (additional wage for working on Saturdays). fi ff ffi fi ffi ff ff ff s fi ff fi fi fi fi 26 - Team composition: - Selection of new hires: - In most cases, the supervisors should be the ones that selects and allocates new hires to teams but sometimes it is the current team the one that is more likely to pick the most suitable person. - The current team, rather than supervisor, should select the applicants for the team if: - The current team already exists for long time. - Applicants are known by the current team. - Team members are engaged in highly specialized tasks beyond the knowledge of the supervisor, so that team members are the only ones that have enough knowledge to do the selection. - Team member selection mechanisms: - Alternating draws: if there are two teams and both team leaders are selecting from the pool they follow an order of preference (the team that makes rst draw is determined by coin toss). Problems: - E ciency of team depends on chance. - Principal-agent problem: the principal might be the rm and the agent would be the team, the preferences of the rm and the team might be di erent. - Bidding for members: in order to attract a team member, the team must give up some of its pro t, which in turn a ects the team members compensation (team has to give up some percentage of its pro ts in order to buy/attract new members). This auctions generally result in an e cient allocation of resources (preferable to alternating draws). Bidding for members solves the previous problems of alternating draws. 8.3 Information Technology (IT) - The business press generally pushes the view that information technology (T) results in decentralization, empowering the workers, breaking up traditional decision making structures and replacing them with more informal structures. - Information technologies (IT) can lead to organizational centralization and narrower jobs. Example of narrow down of job due to IT (Big Data in Mrs. Fields Cookie Co.): - Store speci c data is collected at Mrs Fields headquarter (sales, climate and day of the week). - Based on historical data, computers predict which cookies to prepare - Knowledge from each single stores is used to run all stores more e cient. - IT changes Mr. Fields workforce and business model: store manager’s tasks are reduced and less skilled personnel is able to perform them, which relatively rises the importance of sales skills. Example of narrower jobs due to IT (GPS-tracking of trucks): - GPS-tracking (IT) allows to locate the position of trucks in real time. - Decision making (where next) shifting away from truck driver towards headquarter or computer. - Closer monitoring of truck drivers (driving style, speed and breaks). - IT changes organizational structure of logistic rms: many truck drivers before were independent owner-operators, but now trucks tend to be owned by a company and drivers are employees. - Decentralization tendencies: IT may increase workers knowledge. For instance, IT enables workers to collect and analyze data on their own (workers take some initiative that their supervisor may not have given to them). Availability of email and messengers reduces the degree of control of the supervisor over subordinates. - E ects of IT on job design: Computers as substitutes and complements for people (workers): - Substitutes: rms use computers when there are simple rule based and de ned tasks, when computers are relatively better at performing a task (more reliable, predictable and without motivation problems). Computers are the best options when having simple rule based tasks. - Complements: rms decide to use computers as a tool or as a complement when there are more abstract tasks involved. It mixes computers power and human creativity (abstraction). Computers provide of many facilities and opportunities for workers in many tasks. * Low requirement jobs are tend to be substituted while high requirement jobs are enriched. fi fi fi ffi ffi fi fi ff fi fi fi fi fi ff ffi ff 27 Macroeconomic relevance of IT over time College-high school wage ratio 8.4 High Reliability Organizations (HROs) - High reliability Organizations (HROs): are organizations where the costs of errors are extremely high and where most of the decisions are centralized, except in high pressure periods (where decision making is decentralized). Some examples are hospital emergency rooms or military units during war. - Operations are often tightly interconnected (favors centralized organizations). - Must respond quickly to unforeseen circumstances (favors decentralized organizations). - Structures: common principle of High Reliability Organizations (HROs) is to develop two separate parallel structures: - Quiet periods structure: centralized decision making for normal operations. It’s a preparation for high pressure periods and emphasizes on training (being ready for extreme risks). - High risk periods: decentralized or delegated decision making to lower levels due to the need of immediate execution of the decisions. 28 Topic 9. Performance Evaluation 9.1 Motivation - Human behavior is motivated by two di erent types of motivation: - Intrinsic motivation: rooted in person’s psychology (own values like loyalty or reliability). - Extrinsic motivation: refers to exogenous or environmental parameters (wages or other’s expectations of a person’s performance). The reason why a person is motivated is because receives something in exchange - This motivation distinction matters because it’s easier to alter exogenous patterns (wages or bonuses) than alter the values and views of a person. 9.2 Performance measurement - Performance is a worker’s impact on rm value (the principle want workers to maximize the rm value). Measures or quanti cations of performance: - Broad measures: closely tied to rm value output based measures (stock options / rm value in stock market). - Narrow measures: input and output based measures less tied to rm value (hours working or piece rate, unit). - Problems of performance measurement: - People work in teams, which makes it di cult to lter out the individual impact of a person. - Some workers free ride or shirk, causing asymmetric information. Not everyone knows about the actions of other persons, so the principal cannot perfectly monitor what the agent is doing (some workers are just lucky). - Di culty to measure worker impact on customer satisfaction (quality of customer service). - Measuring performance is costly: - Subjective evaluations for each employee takes time that might be missing for other tasks. - Measurement error may demotivate workers as they don’t feel fairly treated (compensation). People that work harder might receive as much as free riders - Quantitative performance measurement: - Quantitative measures: anything that can be quanti ed (working hours, worker output, customer satisfaction) and can be easily tied to the workers compensation. - Quantitative doesn’t mean objective because a person that woks many hours can produce less output than other in less time, doesn’t make the same contribution to rm value (need to measure real performance). - Properties of “good” quantitative measures: a performance measure should include whatever is controllable by the employee, but exclude whatever is uncontrollable. - Controllable risks: risks that, in case of adverse e ects, the employee can prevent or mitigate some of the damage or, in case of positive events, can prepare for or exploit the opportunity. - For example if a competitor move was predictable and the employee is not able to foresee it even when it was obvious, this employee should be punish cause his performance wasn’t the expected. Other example would be if there’s re in a building and the extinguishers are empty because the employee forgot to ll them even being his responsibility. - Uncontrollable risks: risks the employee cannot change or have in uence on (do not depend on him). - For example, in an economic recession with the stock prices declining, the stock prices would be reduces too, which would reduces the workers compensation even if it is not his fault (there should be no punishment). Other example is a technological advantage of a competitor compared to the rm (reduction in competitiveness and, hence, in revenues). * Compensation should be more closely tied to controllable risks and less to uncontrollable risks. - Distortion: when agents behavior is less aligned with the principles objectives. Distortion arises because of some relevant aspects that are hard to measure (intangibles), due to teamwork (if rm gives individual incentives to the team, team performance high be reduced because there would be no coordination between individuals) and time horizon (historical revenues in the past). - Broad measures (closely tied to rm value) have low distortion but high uncontrollable risk, while narrow measures (less tied to rm value) have high distortion but are less likely to be uncontrollable risky. fi fi fi fl fi fi fi ff fi ff ffi fi fi fi fi fi fi fi 29 ffi fi Paying for Performance the employee must match. If a person has broad measures (closely tied to rm value) has more and di erent rights than a person with narrow measures. - There is a close relationship between the scope of tasks, the performance measure and the decision rights. If an employee is given more tasks or authority, a narrow performance measure will cause greater distortions. Evaluation and job design should match. - Manipulation: arises from the employee’s strategic use of asymmetric information (when information is not known by everyone) after a plan is designed, when maximization of the performance measure does not maximize the rm value (with imperfect monitoring). - For example, if a person is compensated for generating revenues, and there is the option for the customer to return the items, this person can tell people to buy the items and then return them in order to manipulate the revenue (the metric is used with incentive purposes, which degrade its quality as performance measurement). - Broad performance measures tend to be less susceptible to manipulation because the employee would have to change more dimensions of performance in order to manipulate the overall measure. 9.3 Subjective Evaluation - Subjective evaluation: if a supervisor or manager has to do a performance rating, on average, there would be very favorable ratings because they are humans and workers and don’t like to give bad rates to other workers. If this happens it is di cult to select employees to promote cause many have very good ratings but no all of them can be promoted (subjected evaluation is decreasing its use for promotion decisions). - Possible problems: - Grade in ation in universities (students gets better grades for same performance over time). This derives problem because if all the students have the best grades, it’s di cult to make the decision of who to hire. - Favorites problem: a supervisor, manager or teacher mu¡right have his favorites and would give them the best rates even if their performance does not deserve it. - Leniency bias: reluctance to give bad feedback. People don’t want to evaluate low a person and tend to in ate the rating because they are human too and would like to have a good evaluation theirselves. * Subjective evaluations are still important for evaluations made in order to hire, promote or terminate. The issues and problems they own does not make them useless. Subjective evaluation (manager’s responsibility within the destroys of a tornado) - A factory manager’s compensation is based on pro ts and a tornado destroys the (very old) roof of the factory. Should the manager be held responsible (be punished)? - Partial responsibility for partially controllable events: the manager new the roof was old but deferred maintenance. - No responsibility for uncontrollable events: if the manager did not know about the roof, punishments however unfair for the manager, will serve as an incentive to make the factory fully operational as soon as possible. * Performance evaluation should be made thinking on what does the manager know and when he do so. - Bene ts of subjective evaluation: - Complement objected evaluations in order to have a proper risk pro le. - Filter out and properly de ne controllable and uncontrollable events. - Reduce distortions because some dimensions of the job are hard to quantify. As subjected evaluation of the quality become part of the performance measure (compensation of the employee), it would lead to more quality awareness from part of the employee. - Reduce manipulation of the incentive system. If performance measurement is based in a mix measures, it takes much more e ort for a person to manipulate all of them (returns of doing it are not enough). fi ff fi fi ffi fi ff fi ff ffi fl fi 30 fl - Match to job design: to ensure e ectiveness, performance measures need to suit the job. Measures and rights of Topic 10. Rewarding Performance 10.1 Incentives - Firms must use evaluations in order to motivate their employees. - The payment (reward) employees receive is the sum of base salary “a” and a commission rate “b” time performance measure “PM” (the higher the value of the PM, the higher the commission the employee gets): - The value of “a”, base salary is partly determined by the labour market because a rm has to provide a base salary in order to receive applications and hire employees. Base salary “a” is a primarily labor market outcome). - The value of “b”, commission rate or incentive intensity is determined by rms and depends on how strongly relate the performance (hours, units, kms) to the payment (number between 0 and 1). - The value of PM, performance measurement is generated by the worker (employee’s impact on rm’s value) Pay = a + b · PM - If the rm doesn’t implement commission, there would be no incentive to increase e ort and hence, performance. - If the rm implement commission, with every performance increase (e ort), there would be an increase on the pay too. Firm use incentive “b” to exhort e ort from the employee. - Every increment in performance measure (PM) requires more e ort and it gets more and more di cult to increase it. Which plan provides stronger incentives? - Both plans provides same incentives cause are parallel, the di erence is that a2 has a bigger base salary than a1. - Designing compensation plans 1. De ne performance measures (PM): impact of the employee on rm’s value taking into account uncontrollable risks and distortions (incentives of PM less aligned with company objectives) and using subjective evaluation. 2. Incentives (rewards tied to PM / choice of “b”): decide how large the commission rate should be. 3. Choose base salary “a” according to (exogenous) labor market circumstances. Endogenous determination (base salary “a” and incentives “b”) - There are 2 decision makers: a rm (producing personal computers, PC) and a worker (selling PC). - Assumptions: - Firm’s marginal pro t per computer is 1000$ (regardless of total number of PC). - Characteristics of the worker (salesperson): - The more PC the worker sells, the more e ort he has to exert (Ce ort measured in $). - Outside option (wage outside the rm) oopt ≥ 0 (measured in $). - What plan a* and b* would arise in equilibrium? Pay = a* + b* · PM - Key questions: - How many computers would the rm like to sell? - How many computers can the rm (delegating on the agent) sell at most? - The rm has to delegate the selling-task to the worker, who has his own objectives. What does this imply for b* (optimal commission rate? - Given the oopt (worker’s outside option), can the rm determine a*? b* = marginal revenue from increasing PM by 1 unit. Marginal pro t per computer (1000$) = Maximum commission “b” worker would receive. a* > oopt . If oopt = 0, the a < 0 (workers buy the job) fi fi fi fi fi ff ff ffi fi fi fi fi ff fi fi ff fi fi ff fi fi ff 31 oopt a* Q* (PM*) - Buying the job: - Worker cossing between (a1, b1) or (a2,b2) with a1 > a2 and b1 < b2. - Cab drivers: rent a car from a company (buy a job) and earn all the money they get from driving people. - Waiters (US): earning tips (very low base salary) the more e ort the waiter put, the more tips he would receive. - Outsourced sales: sells the products of a company under commission rate (marginal revenue). - Entrepreneurs: chose to start a company knowing that there are low base salary but high commission rate. Job a (base salary) PM b Cab driver rental price for cab Km price for km Waiter working in outside option service quality tips Outsource saler working in outside option sold quantity marginal revenue Entrepreneurs entry ticket for new occupation sold quantity commission rate - Strong incentives tend to give a higher total pay because: - The employee is motivated to work harder. - Attraction of better employees and higher productivity: workers that struggle to increase PM leave the rm, white workers that maximize PM would stay due to the high incentives (b) of going it. (a < 0 and b > 0). - Introduction of risks (performance measured error), that must be compensated with higher a (base salary). - If PM (output) depends on e ort “e” and luck “ε” (random), the volatility of the Pay is increasing in b. - If “ε” (luck) increases, rms reduce b (incentives) and increase a (base salary or risk premium). - The less accurate (precise) the performance measure (PM), the weaker the optimal incentives (b*). PM (Performance Measure) and Distortions - A worker performs 2 tasks: q1 (selling PC) and q2 (customer service). - If the rm wants the worker to put e ort on both activities (selling and customer service), it has to provide the employee incentives for both. Cost function (e ort): Contribution function (quantity): - In order to avoid distortions the rm should make a balance between the and q2 mixing both performance measures. - There are two options: - Implement PM3 plan, which. Includes both tasks for one worker. - If it’s complicate, employee 2 di erent individuals, one for each two tasks q1 Incentives performance measure (PM1, PM2) or tasks (q1, q2). fi ff ff fi ff ff fi ff fi 32 10.2 Paying for Performance Penalty Introducing a threshold as penalty might increase e ort of employees. Employees must work hard cause not doing so would be costly for the company. The scheme motivates employee to put in the e ort in order to reach the threshold in which base salary stabilizes and doesn’t increase with the PM. The salary has to be steep (increase) enough to motivate the worker. (Working very hard doesn’t add any value to the rm so that the salary stabilizes at the point where increase in e ort does not contribute to the rms value. Reward Introducing a threshold as reward insures against bad luck and provides incentive to the employees to increase their e ort. There is a base constant salary in order to insure against bad luck losses of workers. Workers are motivated to work hard in order to reach a higher salary for each increase on e ort. The aim of the reward is to motivate the worker to push the e ort beyond the threshold where every increase in e ort increases rm’s value. Bonus with oor and cap Introducing 2 di erent thresholds, one in the oor and other on the cap incentives the employees to put the e ort on a certain segment of the performance measure. This way the employees are motivated to put e ort on their performance but not too much cause at a point the e ort doesn’t make any contribution to the rm’s value. Lump-sum bonus, demotion or promotion Introducing a threshold that impulses employees to a much higher salary by increasing only one unit of performance or throws back employees that decrease it by one unit would motivate employees to increase performance and don’t reduce it. This happens when a worker is promoted, starts his job with a base salary and when reaches a level of performance and is promoted start earning a much higher base salary. - Implementation of a pay scheme with incentives: - Transparency: share the details of the compensation plan (fair process) and reward good performance. - Gradual transition: expect turnover, convert raises to bonuses (people above threshold must receive bonuses) and provide initial insurance against downside risk (start test cases / role models), communicate and listen. - Reserve the right to adjust over time. - Use 3 stage process to deign plan: evaluate performance, tie pay to performance and set overall pay level. ff fi ff fi ff ff ff fi ff fl ff ff fl ff fi ff ff 33 Topic 11. Career-based Incentives 11.1 Promotions and incentives Mean is the average quantity that relates and combine all quantities of each level. 95th percentile means that 95% of the employees earn under the quantity, and that the 5% remaining earn above it. 5th percentile means that around 5% of the employees earn under that quantity, and that the 95% remaining earn above it. - On the lower levels there is less di erence between pays than in higher levels, all the pays are close to the median (there’s no variance). The spread increases with each level because while moving up in hierarchy. The company starts paying on performance (incentives for e ort or performance that increase rm’s value). - Promotions as incentives: - Person promoted: best performer or person with highest potential performance in the next hierarchical level. - Promotion has two roles (incentive and sorting mechanism) that will be in con ict with each other if the best performer at one level is not the best performer at the next higher level (the best researcher has not necessarily to be the best manager). - A solution to this con ict could be rising the salary of the best performer and leave promotion to the potential best performer in the next hierarchical level. - Unavoidable incentive system: even when the rm don’t want to use promotion as incentive system (increase pay at each hierarchical level to motivate lower level workers), has to increase compensation as employees are promoted. - When a worker gets promoted, creates a strong signal, a signal to potential employers. If current rm doesn’t want to loose promoted worker, must increase salary because outside options are willing to pay more for him. - If current rm knows when worker gets promoted, outside options would increase, it might consider delaying promotion because the rm knows it would have to increase the salary much more. - One way to circumvent outside options is o ering a lifetime employment, which is less constrained by the external market forces. Firms reward loyalty by o ering employment till the retirement of the worker. This motivates workers to stay even knowing outside options give a higher salary. 11.2 Tournaments and standards Promotion rule Tournament Standard Who to promote Person with the highest PM Everyone who exceeds threshold PM* Number of promoted workers Known = 1 Unknown = x Quality of promoted workers Variable quality over time Variable but with lower bound ( > PM* ) Goal Maintain (hierarchical) structure Focus on quality (lower bound) Evaluation Relative (ordinal) evaluation Absolute (cardinal) evaluation Alternative Outside hiring Outside hiring fi fi fl ff ff fi ff ff fi fl fi 34 Absolute vs. Relative performance measurement - Consider the function of performance measure PMD of a person working in Denmark, a function of: - Own e ort (eD): e ort of the sales person. - Local risks (εD): economic conditions in the country (local shocks such as booms or recessions). - Global risks (η): events a ecting the world economy (independent uncontrolled risks). - A rm has to decide between promoting a salesperson in Denmark, or one in Singapore, based on their performance metrics: - Relative Performance measurement: - Relative performance measure does not depend on global shocks (because would a ect both workers). - If the PMD > PMS, relative performance is positive (RP > 0), promote the salesperson from Denmark. - The problem of Relative Performance is that it depends on local shocks, which does nothing to do with the individual e ort of the sales person. Due to local shocks (luck), a person with low e ort can be hired instead of other with more e ort because of having a better economic situation in the country. - When to use promotion based on relative performance (RP): whenever it is less risky than measuring performance directly. Relative performance evaluation is preferable if the global risks variance is higher than the local risks (ση2 > σε2). Performance sabotage - Worker’s A performance is a function of: - E ort towards performance (eP): worker’s own e ort. - E ort towards sabotage (eS): e ort of the worker put into performing sabotage (reduce cooperation). - Local risks (ε): luck or riskiness of a worker. - Global risks (η): events a ecting globally (independent uncontrolled risks). - Relative Performance (RP): - The chance of promoting A increases with the performance e ort and with the sabotage e ort too, while the chances decrease with local risks. If the company can easily detect sabotage, would just re the worker who is putting e ort on it. - This performance sabotage could lead to a very bad scenario in which both workers put no e ort on performance and only engage in sabotaging each other. If they only focus in engaging sabotage and one of them puts more e ort on it, it would be the one promoted. - Solutions for sabotage: - Encourage cooperation by rewarding teamwork between workers (PMA + PMB). - Add subjective evaluations. - Separate competing individuals or teams. - Recruit cooperative workers with values and integrity (who would not sabotage). fi ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff fi ff ff 35 Seniority Pay: W: wage/salary V: worker’s output value (productivity) with high e ort V’: worker’s output value (productivity) with low e ort Alt: value of alternative time use - Seniority is a time based pay, it increases with the time the worker has stayed in the company. The worker stays at the same hierarchical level (doesn’t promote), but the longer he stays in the rm, the more salary the worker would get over time. - When a person has been recently hired, the value of the salary is lower than the value generated for the rm (productivity), In early periods, workers receive less money than they generate (buying the job). - The worker will have a lower wage than value he generates till a threshold in time where he would start earning a bigger salary than the value he adds to the company. - There is a point in the worker’s lifetime (old) when starts getting less productive and being less motivated so that the value generated for the rm drops. - This pay method encourage workers to stay at the rm at least till the period when they would earn at least the same quantity they generate for the rm (salary = value produced). Low productive workers are red. fi fi fi fi fi ff ff fi 36 Topic 12. Options and Executive Pay 12.1 Employee Stock Options (ESO) - Call stock option: right to purchase/buy company stock (at some point in the future) at predetermined exercise price K. If in the moment of the purchase the exercise price is smaller than the stock price (K < S), the worker (stockholder) can make intrinsic value pro t (S - K). Calling (buying) of the stock ( stock goes to person) - Put stock option: right to sell company stock at a predetermined price K. It can pro t from the selling if K > S. - Employee Stock Options (ESOs): employee options are always calls (buying the stock). Stock options motivate employees because they would earn pro t if the value of the rm increases. - Employee options cannot be treated (sold), only exercised (exercise rights usually after 3-5 years). The options are tied to the employees, and if the worker leaves the rm, all unexercised options are usually lost. - Reasons to o er Employee Stock Options: - Source of rm nancing: to have money for other purposes than paying high salaries (useful for start-ups). - Employee commitment: if the workers have stock options, they might be more involved with the company’s objectives and in maximizing the rm’s value cause their pro tability depends of it. - Reduced turnover, in particular if employees cannot exercise their options anymore when leaving the rm. * Economic theories suggests that there are better ways of nancing a rm, increasing employee’s commitment and reducing turnover (unsupported by economics). - Incentive e ects of options: - Options are “good” performance measures (PM) because are similar to stock. Options are very risky to lower level employees because lower level workers have little impact on the stock price compare to higher levels. - Options can provide stronger incentives than stocks (in higher level employees): - Options only pay o if the di erence between the stock price and the exercise price is positive (S - K > 0). - A share of stock is more valuable than an option because it can be sold and make more pro t. - The rm can issue a greater number of options than of stocks (because options have less value). - Because there’s a larger number of options, options can motivate workers more to increase e ort. - If stock prices fall, then options do not motivate at all (options are worthless). 12.2 Executive (CEO) Pay E ect of CEO e ort (e) on compensation: - CEO puts e ort into the rm so that if e ort increases, stock value increases too, and if stock value increases, the compensation or reward of the CEO increases too. E ect of rm size on executive incentives: - The impact of e ort on the stock value tends to increase with the rm size. - The impact of stock value on compensation tends to decrease with the rm size. * As both impacts compensate each other, one decrease and other increase, rm size doesn’t have any e ect on executive incentives. - Other incentives that a ect executive (CEO) behaviors: - Pressure by stakeholders: help to increase or reduce rm value depending on they information they provide. - Product market competition: if the rm has to constantly increase the quality of the product and reduce the prices in order to stay competitive, CEO won’t have a too high compensation because the rm needs money. - Market for corporate control: bad decisions made by CEO make stock value decrease (CEO hostile takeover) - Oversight by the Board of Directors: are able to change decisions made by the CEO. fi ff fi fi fi fi fi fi fi fi fi fi fi ff fi fi fi ff fi fi fi ff ff ff fi ff ff ff ff fi ff fi fi ff ff 37 - Governance (a ect executive behavior without monetary incentives): - Board composition: should be formed by outsiders (not insiders, CEO neither relatives) and might limit outsider CEOs, which may be too sympathetic (might compensate other CEOS). - Compensation Committee: should not be formed by insiders, take an active role in designing the package and monitor compensation consultants for con icts of interest. - Process: meetings with and without the CEO and provide Broad of Directors with data if the employees. - Culture: people are open and independent, roles and responsibilities of members and committees are clearly de ned, and people talk about values and the proper/correct of doing things. fl ff fi 38