Treat employment as a relationship with outside options, incomplete contracts, and institutions.
By the end: Explain how productivity, worker alternatives, and bargaining affect pay without assuming wages measure worth.
Labor is not an ordinary stored commodity
A worker sells labor services under a contract, not ownership of their person. Effort, safety, scheduling, discretion, and future progression are difficult to specify completely. Employers and workers have different information and outside options. Search costs, transport, care responsibilities, discrimination, and immigration restrictions can limit the alternatives available to a worker.
In a competitive benchmark, a firm compares the revenue contributed by an additional worker with the extra cost of employment. This is a conditional demand rule, not a statement that every worker's observed wage equals their social contribution. Revenue depends on product prices, teamwork, equipment, demand, and market power. Important unpaid work can have no market wage at all.
Compare the value of a job with the cost of taking it
An hourly wage is a payment for an hour of paid labor, not necessarily an hour of the worker's total committed time. Commuting, searching, maintaining availability, and arranging care can affect whether the job improves the worker's available options. Some costs are monetary, while others use time or impose risk. An outside option is the best feasible alternative if the agreement fails. It is not whatever wage a worker wishes to receive, nor necessarily a job available immediately. To compare it with a job, put the two alternatives on matching time and cost bases.
Consider a separate paper case for one day. A job pays 18 per hour for eight hours, giving gross earnings of 144. Travel costs 12 and additional care costs 28, leaving 104 after these job-specific expenses. The job also requires two hours of commuting. Over ten committed hours, that net receipt is 10.40 per hour. That figure is not a revised legal wage rate. It is a comparison of money left after specified costs with the worker's total committed time. Taxes, benefits, safety, fatigue, and future prospects remain outside this deliberately narrow calculation.
Suppose a nearby job pays 15 for eight hours, with no additional travel or care bill and no commute in this simplified comparison. It leaves 120 over eight hours. Its lower posted wage delivers 16 more money per day and uses two fewer hours. This does not show that workers always prefer nearby jobs. The distant job might provide training or more secure future hours. It shows why posted wages alone do not rank feasible offers. A care subsidy or transport service would change the comparison through a specific budget item rather than by changing the worker's underlying preferences.
A lower posted wage can leave more daily cash
The invented distant job leaves 144 minus 12 travel and 28 care, or 104 per day. The nearby offer leaves 120, despite its lower hourly wage. These bars compare daily money, not hourly pay: commuting also raises distant-job time from eight to ten hours. Training, taxes, benefits, and security could reverse the ranking but are outside this calculation.
The employer's calculation has a different boundary. A team may produce sales worth 1,000 in a day, but those sales are not automatically the contribution of one additional worker. The employer must compare output with and without the hire, allowing for equipment, materials, and other workers. If adding a worker raises receipts by 160 while adding 40 of nonlabor costs, the net contribution available to cover that worker's compensation is 120. An offer above 120 reduces the employer's payoff in this one-period comparison. Training benefits or future sales could change the relevant horizon.
Neither comparison measures a person's moral worth. The worker's reservation terms depend on actual opportunities and obligations. The employer's revenue depends on customers' budgets and the product market. A public service can benefit people who cannot pay enough to generate comparable commercial revenue. Conversely, high revenue can include gains obtained through market power. Explaining how a wage can be sustained in an employer's budget is different from judging the service's social value. Keeping those questions separate also prevents interpreting a constrained acceptance of work as evidence that every aspect of its terms was freely preferred.
Extra receipts are not all available for pay
For the invented additional hire, compare receipts with and without that worker, not the team's total sales. Subtracting 40 of additional nonlabor costs from 160 of added receipts leaves a compensation ceiling of 120 in this one-period budget. Future training benefits could change that ceiling; neither this residual nor the worker's alternative measures social worth.
Separate the feasible surplus from its division
Our bargaining model gives a job a net revenue contribution of 30 per hour after other costs. The worker's outside option is equivalent to 10 per hour. There is a potential surplus of 20 relative to that alternative. If the worker receives one quarter of that surplus, pay is 15 and the employer retains 15. A better outside option of 18 raises pay to 21 at the same bargaining share.
This formula illustrates bargaining; it does not estimate a country's wage-setting process or employment response. Unions, job mobility, benefits, labor law, and workplace organization can affect bargaining and outside options. Productivity can enlarge the available surplus without dictating how it is divided. A larger pie and a larger worker share are different changes.
Derive the bargaining interval before choosing a share
The bargaining model starts with a feasible job, not an equation estimating labor demand. Net contribution is the money available from the job after other costs. With contribution fixed at 30 per hour, a wage of 22 leaves the employer 8. A wage of 28 leaves 2. An employer with an alternative normalized to zero can accept either without a negative residual in this model. A worker with an outside option of 10 gains 12 at the first wage and 18 at the second. Both bargains generate the same total gain relative to disagreement.
That gain is 20 per hour because employing the worker creates 30 while giving up an alternative worth 10. The feasible wage interval runs from 10 to 30. At the lower endpoint the worker receives no gain above their alternative. At the upper endpoint the employer receives no gain above its zero alternative. Outcomes inside the interval distribute the same gain differently. The sharing parameter picks a point within that interval. Calling it bargaining power summarizes the division assumed in the exercise; it does not independently explain the process that produced the division.
Different wages divide the same job contribution
Both invented bargains allocate a contribution of 30. The worker's first 10 replaces the forgone outside option; only the remaining wage is a gain from agreement. Raising pay from 22 to 28 transfers 6 from employer residual to worker gain without enlarging their combined surplus of 20. The employer's alternative is zero; different alternatives would change the bargaining interval.
For a paper calculation within the widget's available shares, give the worker 40% of the surplus. Forty percent of 20 is 8, so the wage is the outside option of 10 plus 8, or 18. The employer keeps 12. Check conservation by adding the worker's gain of 8 to the employer's 12. Their combined gain is 20, not 30, because the worker forgoes the alternative worth 10. Adding the full wage of 18 to the employer's residual instead gives the job contribution of 30. The two totals answer different questions.
Now keep the 40% share but raise the outside option to 18 using the better-alternative setting and then adjusting its share. The total gain falls to 12. The worker receives 4.80 of that gain, so pay becomes 22.80 and the employer keeps 7.20. The wage rises by 4.80 rather than the full 8 increase in the outside option. The worker gives up a more valuable alternative, leaving less surplus to divide. This is conditional arithmetic, not evidence that a particular benefit policy changes wages by this amount.
The employer's alternative is also consequential. As a paper extension outside the widget, suppose contribution remains 30, the worker alternative is 10, and the employer can obtain a payoff of 6 elsewhere. Agreement now creates only 14 above their combined alternatives of 16. Giving the worker 40% of that 14 produces a wage of 15.60. The employer receives 14.40, of which 6 replaces its alternative and 8.40 is its gain. The actual widget keeps the employer alternative at zero. Altering it on paper shows why identifying both parties' real alternatives matters before interpreting an observed wage.
Work through the mechanism
An illustrative surplus-sharing rule
w₀ is the outside option, v the job's net contribution, and β the worker's share of the surplus. Money values are per hour; β is a fraction between zero and one, not an estimated parameter.
Separate productivity from the division of surplus
Net contribution is held at 30 currency units per hour after other costs. The worker's outside option is a comparable hourly value; the employer's alternative surplus is normalized to zero. Changing the share or preset replaces the hypothetical bargain. No employment quantity, wage-floor effect, or behavioral estimate is calculated.
Bargaining case
Outside option
Worker share of surplus
Wage
Employer residual
Baseline
10
25%
15
15
Better alternatives
18
25%
21
9
Stronger worker bargaining
10
50%
20
10
Job contribution is held at 30 per hour. This arithmetic does not model job creation, legal wage floors, or the determinants of bargaining shares.
Use the wage controls to distinguish two changes
The wage lab has three named presets and one slider for the worker share of surplus. Its contribution stays at 30 per hour. Start with Baseline alternatives, which sets the outside option to 10 and the share to 25%. Read the outside-option marker separately from the full wage bar. The part of pay below that marker compensates for giving up the worker's alternative. Only the part above it is the worker's gain from this agreement. The employer bar shows its whole residual because its alternative is held at zero.
Move the share to 60%, keeping the baseline outside option. The surplus remains 20. Sixty percent gives the worker 12 above the alternative, so pay becomes 22 and the employer residual becomes 8. At a 65% share, the worker receives 13 above the alternative, giving a wage of 23 and residual of 7. A five percentage point share change transfers 1 per hour because five percent of the 20 surplus is 1. It does not add five percent of the existing wage. The denominator for the share is the surplus above disagreement.
Select Better outside option. This replaces the outside option with 18 and the share with 25%, even if the slider was previously at 65%. It is a complete hypothetical bargain, not a cumulative improvement added to the previous slider state. To isolate an outside-option change at a common share, set 60% again. The surplus is now 12; the worker gains 7.20, pay is 25.20, and the employer keeps 4.80. Compared with baseline alternatives at 60%, the outside-option increase raises pay by 3.20. Keeping the share fixed makes this comparison interpretable.
Hold the share fixed to isolate the outside option
At a common 60% worker share, the invented outside-option increase from 10 to 18 raises pay from 22 to 25.20. The increase is 3.20, not 8, because the surplus available for sharing contracts from 20 to 12. Contribution stays 30 and the employer's alternative stays zero. Changing the share simultaneously would mix two mechanisms; these values are not policy estimates.
The slider runs from zero to one in steps of 0.05, meaning zero to 100% of the surplus. At zero, pay equals the outside option. At one, pay equals the contribution, leaving the employer zero. These endpoints are useful checks on the calculation, not predicted stable employment arrangements. Stronger worker bargaining selects an outside option of 10 and a share of 50%. Selecting Baseline alternatives restores its initial assumptions. There is no separate playback or accumulated employment history to reset. Controls replace the displayed bargain immediately, and the lab contains no employment quantity or contribution slider.
Changes in productivity must therefore be explored as paper cases rather than through an invented control. If a better machine also requires a higher equipment payment, gross sales gains cannot all count as increased net contribution. First subtract the additional nonlabor costs, then apply a specified bargaining rule. A union contract, wage ladder, or statutory floor may replace that rule altogether. The display helps distinguish a larger feasible surplus from a different division, but cannot establish how actual institutions choose either one. To infer those mechanisms, investigate contracts, alternatives, and job outcomes rather than treating the chosen share as an observed causal parameter.
A machine gain and a wage rule are separate inputs
Paper extensions beyond the wage controls
Comparison
Hold separate
Wage implication
Machine raises sales
Extra equipment cost
Net gain still unknown
Same sharing rule
Outside option fixed
Divide net gain
Union contract
Contract pay terms
Share rule may not fit
Statutory floor
Legal minimum pay
Check rule against floor
The wage widget fixes net contribution, so a machine improvement is a paper extension, not an available control. First account for its additional equipment payment. Only then can a surplus-sharing rule divide the remaining gain. A contract or floor may replace that rule; without its terms, neither higher gross receipts nor a chosen share predicts the resulting wage.
Minimum wages and employment need a model and evidence
If employers have wage-setting power, sometimes called monopsony power, attracting more workers can require increasing pay. A moderate wage floor can then raise pay and, over some range, employment relative to an underpaying benchmark. In a competitive model, a binding wage floor can instead reduce labor demand. A sufficiently high floor can reduce employment in either model. Coverage, enforcement, hours, prices, and firm entry also matter.
One monopsony mechanism resembles the uniform-price problem in reverse. Hiring another worker may require raising pay for existing workers too, so marginal hiring cost can exceed the wage. A moderate binding floor can make additional hires available at the same wage over a range instead of requiring repeated increases for the whole workforce. Marginal hiring cost can then fall even while the paid wage rises. This mechanism requires the relevant wage-setting and labor-supply conditions; it is not a general prediction for every employer.
These theoretical possibilities do not settle the effect of a particular policy. Natural experiments compare similar affected and unaffected settings under explicit assumptions. Even a credible estimate may apply mainly to the workers, industries, and range of wages studied. Distributional judgments must also account for who gains pay, who loses hours or work, and how costs are passed on.
Why a wage floor can change marginal hiring cost
A separate paper employer illustrates monopsony without pretending the bargaining widget simulates it. Assume workers are identical, each contributes 40 per day after nonlabor costs, and all employees must receive the same daily wage. One worker is available at 10, two at 15 each, three at 20 each, and four at 25 each. These are labor-supply offers to this employer, not measured data. Total payroll at those employment levels is 10, 30, 60, and 100. The cost of moving from one to two workers is therefore 20, even though the second worker's wage is only 15.
The extra 20 consists of 15 paid to the additional worker and a 5 raise for the first. Moving from two to three raises payroll by 30. Moving from three to four raises it by 40. Employer residuals at one through four workers are 30, 50, 60, and 60, after subtracting payroll from contributions of 40, 80, 120, and 160. Zero workers yields zero. With a stated tie rule choosing fewer workers, the employer chooses three and pays 20. The tie rule belongs to this paper example, not the wage lab.
Introduce a binding daily wage floor of 25 and suppose four workers are willing to work at that wage. Payroll for one through four becomes 25, 50, 75, and 100. Residuals become 15, 30, 45, and 60. Four workers now uniquely maximize the employer's residual. The wage rises from 20 to 25 while employment rises from three to four. The floor removes the need to raise existing workers' wages again as employment expands up to four. This is a conditional example of a mechanism, not a general law that higher wage floors increase hiring.
A wage floor changes which workforce pays best
For this invented employer, each worker contributes 40 daily. Without a floor, three and four workers both leave 60, so the stated fewer-workers tie rule selects three. At a wage floor of 25, three leave 45 and four leave 60, making four uniquely best. Smaller workforces yield lower residuals in both cases; a higher floor need not preserve this result.
Contrast a floor of 45 in the same capacity-limited paper business. Each worker contributes only 40, so every hire now reduces the employer residual by 5. Zero workers is best. A competitive employer facing a fixed wage also lacks the original rising marginal payroll mechanism. Thus employment responses depend on wage-setting conditions and the policy's level. Hours, task changes, product prices, and exit are omitted here. Adding them could alter both measured employment and who bears the cost. The person legally paying wages need not bear the entire economic cost if product prices or other payments adjust.
An empirical study must identify a comparison outcome without the wage-floor change. Before-and-after hiring alone mixes the policy with demand shocks, technological changes, and workforce composition. A comparison group helps only if it represents the affected group's credible untreated path and is not itself materially affected. Even a well-identified average effect can combine gains for some workers with losses for others. Reporting employment counts without hours or earnings can miss that distribution. Deciding whether a policy is desirable then requires explicit judgments about those outcomes; neither the bargaining arithmetic nor this monopsony example supplies those judgments.
Exercises
Productivity need not go entirely to wages
Keep the outside option at 10 and the worker share at 25%. Net contribution rises from 30 to 50 per hour. What happens to wage and employer residual?
Hint
Apply the share to the surplus above the outside option.
Worked answer
Contribution
Wage
Employer residual
30
15
15
50
20
30
The model separates a productivity gain from the rule for distributing it.
The surplus rises from 20 to 40. Pay rises from 15 to 20, while the employer residual rises from 15 to 30. Of the 20 gain, the worker receives 5 under these assumptions. A different bargaining arrangement could divide it differently.
A wage is not a social-value meter
A care worker earns less than a financial trader. What additional facts would be needed to compare their social contributions?
Hint
Market revenue and social benefit are not the same quantity.
Worked answer
Observed
Not established by observation
Different wages
Different social value in the same ranking
Different funding and demand
Different intensity of human need
Different outside options
Different effort or moral deservingness
An explanation of wage formation is not a justification of every wage difference.
You would need evidence on the services produced, benefits and harms to others, scarcity, institutions, and alternatives. Wages also reflect funding arrangements, bargaining, and market demand backed by income. The pay ranking alone cannot establish a ranking of social worth.
Sources and further reading
The lessons and answers stand alone. These optional references identify underlying models, evidence, and perspectives. All numerical examples here are invented unless explicitly attributed; none is a forecast or a personal investment recommendation.
Ownership, organizational authority, incomplete employment contracts, and worker incentives. Our wage-sharing formula is an illustrative bargaining rule, not an estimated wage equation.