MRP breaks neatly into two pieces:
MRP = Marginal Physical Product (MPP) × Price of Output
Marginal physical product is how many additional units of stuff one more worker makes per hour. Price of output is what each unit sells for. Multiply them and you get the dollars one more worker's hour adds to revenue. That number is the ceiling on what a rational employer will pay that worker.
Run the numbers
Picture a small bakery. A new baker can produce 30 extra loaves per hour. Each loaf sells for $5. The baker's MRP is straightforward:
MRP = 30 loaves × $5 = $150 per hour
That $150 is the most the bakery could pay this baker before the hire stops making sense — but it is not the wage. The wage is set by what it takes to attract a baker in the labor market, and skilled bakers are not so scarce that they capture the full $150. Say the going rate for bakers is $22/hour. The bakery earns the $128 gap as part of its return on capital, brand, ovens, and location. The baker earns $22 because that is what the next baker would accept.
Now change one variable at a time and watch the wage move.
Raise the price of output. Suppose the bakery rebrands as an artisan shop and each loaf now sells for $9. The baker's physical output is unchanged at 30 loaves, but MRP jumps to 30 × $9 = $270/hour. The same hands, the same loaves — but attached to higher-priced output, the work is worth more, and in a competitive market the wage for that baker drifts upward. This is why identical skills pay more at higher-margin employers: an accountant at a hedge fund and an accountant at a charity do similar work, but the output their work attaches to sells for vastly different sums.
Raise productivity. Now give the baker a better oven that lets them produce 50 loaves/hour at the original $5 price. MRP = 50 × $5 = $250/hour. The lesson is that capital makes labor more productive, and more-productive labor commands higher pay. This is the engine behind the long-run link between productivity growth and wage growth that the Bureau of Labor Statistics tracks in its productivity program. Workers in capital-rich economies earn more not because they are better people than workers elsewhere, but because each hour of their labor is multiplied by more and better tools.
Productivity sets the ceiling; scarcity sets the rent
MRP explains the maximum a job can pay. But two workers with the same MRP can earn very different wages, and the reason is scarcity. Productivity determines the size of the pie a worker can help bake; scarcity determines how much of that pie the worker can keep versus how much goes to the employer.
Consider why surgeons out-earn nurses even though both are essential and both are productive. According to BLS Occupational Employment and Wage Statistics, the median annual wage across all occupations was about $48,060 in May 2023, registered nurses earned well above that, and surgeons earned multiples more. Surgeons attach to extremely high-value output (life-saving operations the hospital bills heavily for), so their MRP is high — but the decisive factor is scarcity. The pipeline to become a surgeon is brutal: a decade of training, fierce selection, and licensing barriers. That artificial and natural scarcity means hospitals must pay near the top of the MRP range to attract one. Nursing is demanding too, but the pool of qualified nurses is far larger, so employers can fill roles without bidding wages to the ceiling.
Scarcity is the lever, which is why the practical advice that flows from this framework is "develop skills that are both high-value and hard to replace." A skill that raises your MRP but that millions of others also have will lift the whole market's wage, not yours specifically. A skill that is both productive and rare is what lets you capture the gap.
Where the framework bends
MRP is a powerful baseline, but treating it as an iron law overstates the case. Real wages diverge from MRP for reasons worth naming honestly.
Bargaining power shifts the split. When workers organize or when labor is tight, they capture more of the gap between MRP and the market wage; when employers hold the leverage, they capture more. A single dominant employer in a town can pay below MRP because workers have nowhere else to go — the monopsony case the Richmond Fed examines in its research on measuring employers' market power. Unions, minimum wages, and tight labor markets all push in the other direction.
Information is imperfect. Employers cannot perfectly measure any individual's MRP, so they rely on proxies — credentials, experience, prior salary. Those proxies are noisy, which is part of why negotiation and job-switching move pay: you are correcting the market's estimate of your MRP.
Discrimination drives a wedge. When workers of equal productivity are paid unequally based on group identity, the gap is a real-world departure from the MRP prediction — one the framework itself flags as inefficient, since underpaying productive workers leaves profit on the table for employers willing to hire them.
What to do with this
The MRP lens turns vague career advice into something operational. To raise your wage, you have exactly three levers, and they map onto the formula. Raise your marginal physical product by getting more done per hour — through skill, tools, and focus. Attach your work to higher-priced output by moving toward industries, firms, and roles where what you help produce sells for more. And make your particular bundle of skills scarcer relative to demand, so you capture more of the gap between what you produce and what the next person would accept.
None of this requires believing the market is fair. It requires understanding what the market is actually paying for — not your hours, not your needs, but the revenue your hours generate and how hard you are to replace. That is the uncomfortable, useful truth the MRP framework hands you.
