Personnel economics: why firms design jobs, pay, and promotion the way they do — from the simplest spot market for labor up through incentive pay, unions, and the legal boundaries around all of it. Most of this course is conceptual rather than computational, so I've added calculators only where a real formula sits behind the idea — from 33032: Managing the Workplace.
The baseline case, before anything about incentives, training, or long-term relationships enters the picture: a frictionless labor market where firms simply pay each worker what they're worth.
The value a worker adds to the firm's revenue at the margin. In the spot market, this equals the wage — workers are paid exactly what they're worth, no more and no less.
Marginal productivity falls as more workers are hired — the reason labor demand curves slope downward.
The total value created by a match: the firm's surplus (MRP − wage) plus the worker's surplus (wage − reservation wage). A bigger total surplus means a more economically efficient match, however it's split between the two sides.
Notice the wage cancels out of the total — it only determines how the surplus is split, not how much surplus exists in the first place. That's the efficiency/distribution distinction that runs through the whole course: policies can change who gets the surplus (distributional effects) separately from whether they grow or shrink it (efficiency effects).
Raise the wage and two opposing forces pull on how much someone works — which one dominates isn't obvious, and it shapes how policy affects hours worked.
A higher wage raises the reward for working, so workers give up leisure for more hours — pulling hours up.
A higher wage lets workers hit their target income in fewer hours, so they work less — pulling hours down.
Who gains and who loses from a policy — the incidence depends on the relative elasticities of labor supply and demand, exactly like tax incidence.
Separately from who wins or loses, does the policy grow or shrink total surplus?
The fraction of a worker's prior wage they receive while on benefits. A higher ratio weakens the incentive to return to work — this is the classic moral-hazard trade-off behind unemployment insurance design.
Once workers differ in ways firms can't fully observe or price, the simple spot market starts to break down — and firms respond by building longer-term relationships instead.
People vary on some attribute the firm can't price — either because it's unobservable or because pricing on it is illegal — and the gap between the marginal and average person in the pool can be large.
People tend to find jobs that suit them the longer they stay in the labor market, as both sides learn more about the fit.
The intellectual capital an individual holds — best thought of like any other form of capital. Firms want to induce employees to invest in it, whether through formal education (general training) or on-the-job programs (specific training).
Workers value income security; firms can absorb income risk more cheaply than individual workers can, which shapes how pay is structured.
Workers sometimes acquire costly, visible skills not because those skills raise productivity, but simply to show the market how capable they are.
Because relationships between workers and firms last a long time, firms have tools beyond the wage itself to keep people motivated and honest.
The firm offers a contract whose terms are attractive only to certain types of workers — letting the contract itself sort the applicant pool.
Because it often takes time to see the true effect of someone's actions, it can make sense to defer compensation until the results actually come in.
Paying above the reservation wage — what a worker could get elsewhere — on purpose. Reasons include: a better applicant pool, reciprocity (extrinsic pay reinforcing intrinsic effort), and simply making the job more valuable than the next best alternative, which discourages shirking.
Every compensation scheme is a set of design choices — what to measure, how precisely, and at what level (individual vs. group) — each with its own failure modes.
Marginal: how much the agent gains from one more dollar of output at the current margin. Average: what fraction of total output is paid to the agent. The two can differ a lot, and it's the marginal incentive that actually drives behavior.
Returns an agent captures simply from knowing more than the person writing the contract does.
Judging a worker against peers strips out common shocks — but it can invite collusion between the people being compared, and can shift a manager's focus toward beating a rival rather than performing well in absolute terms.
Even a well-designed pay scheme depends on who's watching, how the reward is framed, and what actually drives people to put in effort.
The same expected pay can motivate very differently depending on whether it's framed as a bonus to be earned or a penalty to avoid — loss framing tends to hit harder.
Doing an activity for its own inherent satisfaction rather than for an external reward (e.g., charitable work) — and heavy-handed extrinsic incentives can sometimes crowd it out.
People tend to perform worse in high-stakes, adrenaline-charged moments — though this effect is much weaker when the stakes are simply "this is my job" rather than a one-off high-pressure event.
In teams, a worker under-provides effort because they have to share the extra output with everyone else. Solutions include paying more on individual output and making wages very sensitive to output.
A key role of profit-sharing isn't the direct incentive — it's that it causes workers to start monitoring each other.
Sounds odd, but it can help: it shares risk (outside opportunities are often correlated with the measure), and it's a benchmark the worker can't game or ratchet up over time.
Money isn't the only lever — career concerns and the promise of a better job later can motivate just as much, but they come with their own credibility problems.
Unwritten, moral-obligation-based promises that substitute for a formal contract. Their weakness: if pay is tied to a subjective evaluation, bosses have an incentive to lie about performance to avoid paying out.
Collective bargaining changes both the level of pay and how tightly it's tied to individual performance.
Under the Taft-Hartley Act (1947): workers get the same representation rights without being required to pay union dues.
A Democratic proposal that would have added: employee punitive damages, immediate unionization with no waiting period, and an arbiter to set contract terms after a period without agreement.
A strike during an active contract — normally prohibited, except over genuine safety issues.
"Employment at will" is the default — either side can end the relationship at any time — but a handful of legal exceptions and anti-discrimination statutes carve out limits on that default.
Sex, religion, race, age (40+), and disability — covering hiring, firing, promotion, and job conditions.
Disparate impact: harming a protected group without intent. Disparate treatment: purposely harming someone based on a protected category.
Reasonable accommodation: could a simple change have avoided the problem? Business necessity: the job genuinely cannot be performed without the requirement in question — the "minimum necessary standard."
Employers are always liable for their employees' discriminatory conduct, and sometimes for customers'. Disability covers conditions that affect major life activities; pregnant employees must be treated the same as any other temporarily disabled employee.
Two competing views on when to tie pay to performance, plus the classic economic logic behind what actually deters bad behavior.
Use pay-for-performance in low-uncertainty environments — in a highly uncertain one, imposing risk on workers raises compensation costs (the risk premium they demand).
Use pay-for-performance in high-risk, high-autonomy environments instead — when the firm doesn't know exactly what workers should be doing (classically, start-ups), strong incentives are what get workers to use their own initiative well.
The standard economic logic of deterrence (Becker): someone weighs the guaranteed benefit of misbehaving — taking a bribe, cutting a corner — against the expected cost of getting caught, which is just the probability of detection times the penalty if caught. Raise either the probability of detection or the penalty enough, and the expected cost outweighs the benefit, deterring the behavior even though enforcement itself is imperfect.
Enter the benefit someone gets from misbehaving, the probability they're caught, and the penalty if caught.