Dr. Taylor slid the envelope across the conference table in October. Inside was her productivity bonus statement, and she had already done the arithmetic out loud twice. Her collections for the trailing year: $462,000, the largest book any non-partner at Acme Pediatrics had ever carried. Her bonus, under a contract formula paying 20 percent of collections minus allocated practice expenses: $1,364. “I produced almost half a million dollars,” she said, evenly, “and the formula says my share of the upside is a car payment. Explain the $455,000 of ‘my’ expenses, or I’m going to let the hospital recruiter buy me dinner.”
The partners could not explain it on the spot, and that was the larger problem. The expense allocation had been set by a former administrator using a percentage nobody could reconstruct, under a contract permitting the practice to set it “in its discretion.” The bonus was legal and arithmetically correct. It was also organizationally indefensible, which is a separate and more expensive category. The rule of thumb PMI hands every practice is blunt: any compensation formula a smart physician cannot recompute from the monthly package within 30 minutes is a resignation letter in escrow.
Physician compensation is the single most common source of partner conflict, and the field has settled on five basic models. Survey data on physician practices put straight salary in roughly 21 percent of practices, salary with bonuses in 31 percent, a deliberate blend of salary and productivity in 24 percent, pure productivity in about 6 percent, and custom arrangements in the remaining 18 percent. Prevalence is the least interesting fact about that list. The useful questions are what each model rewards and what each model breaks.
Straight salary buys stability and cooperative behavior: nobody guards patients from colleagues, and nobody resents the unprofitable work of phone calls and forms. What it breaks under is effort drift—identical paychecks paying diverging producers, which is exactly the pattern that had been corroding Acme’s own partnership upstairs while Dr. Taylor’s envelope went wrong downstairs. Salary with a bonus is the field’s modal answer, and its integrity depends entirely on the bonus being real. A bonus that pays regardless of performance is a wage increase on a delay. A bonus that never pays is a morale subtraction sold as an incentive, which is precisely what $1,364 was. The salary-productivity blend pays a base for a defined floor of work and a productivity payment above it; it is the right answer for most stable groups, and its weakness sits out in the open, which is complexity. Blends require measurement, arithmetic, and annual maintenance, so they succeed in practices with governance and decay in practices without it. Pure productivity is precise about volume and blind to everything else—citizenship work, complex panels, payer mix, and the mentoring of the next hire. The fifth model, the custom tiered architecture, is what real pediatric practices most often assemble: productivity-based pay for physicians who want it, salary plus bonus for nurse practitioners, straight employment for therapists and ancillary clinicians.
Work relative value units are the currency that makes any of this comparable across practices, because they measure physician work independent of what a payer happened to pay for it. The reference stack is short enough to carry in the head. Compensation per work relative value unit runs $35 to $50 for general pediatrics. Compensation-to-collections ratios of 25 to 35 percent mark the conventional healthy band, with 20 to 40 percent the boundary of defensible. Production benchmarks put the national median near 5,600 units per full-time equivalent, with the 10th-to-90th range running 3,200 to 8,800. And collections per unit—what the practice’s own payer book actually converts work into—ran $50 at the 25th percentile, $67 at the median, and $98 at the 75th in 2025 pediatric data, against Medicare’s roughly $70.
Those figures earn their keep as a three-way check, and one worked pass teaches the method. A proposal puts $166,000 of compensation against 3,680 work relative value units and $415,000 of attributed collections. The per-unit check passes without comment at $45.11. The ratio check strains at 40 percent, outside the healthy band and at the edge of defensibility. Two checks disagreeing is the method working, and the third number locates the disagreement: collections per unit compute to $113, roughly 15 percent above the top of the plausible conversion band. No pediatric payer book converts work into dollars at that rate, so the data are wrong rather than the deal. In the engagement behind that example the cause was annualization—the physician had worked eight months, her true run rate was about 5,520 units, and corrected, the package computed to $30.07 per unit, below band. The negotiation that followed was about raising her pay rather than restraining it, the opposite of what the uncorrected ratio implied.
A fourth gate sits apart from those three, and it is the one Dr. Taylor’s contract never had. The three checks ask whether the price is right. Affordability asks whether this practice can fund it. The governing inequality admits no exceptions: overhead percent plus compensation percent cannot exceed 100 percent of collections. Break-even revenue equals total loaded provider cost divided by one minus the overhead rate, and the productivity threshold is break-even raised by whatever margin the practice decides to keep. Budget the bonus backward from there—decide what the practice must keep, find the collections level that delivers it, and share cleanly above that line—rather than picking a rate that sounds motivating and discovering its cost at year-end.
Overhead allocation is where formulas go to die, and the repair is a single principle: allocate from the practice’s published overhead rate, applied identically to every clinician, computed on a stated convention. Discretionary percentages fail the conference-room test eventually, always at the worst moment. Quality and value incentives sit on top of the productivity layer and need the same discipline—each metric defined, its data source named, its measurement period stated, and its calculation written before the year begins rather than assessed after it ends.
Then decide where the formula lives, because that decision is governance rather than arithmetic. The compensation methodology, and the vote required to change it, belong in the partnership agreement at the supermajority tier, so no future majority can reprice a colleague’s economics on a simple vote. Each partner’s individual numbers belong in an employment agreement, which is the subject of article 8. Build an annual review onto the calendar rather than onto grievances, give every partner the right to audit the formula’s inputs, and confirm that the design survives the departure or addition of any single partner. The companion textbook, Pediatric Practice Management: The Fundamentals, carries the full model with its worked thresholds for practices building one from scratch.
Red flags. A compensation formula that exists only in a spreadsheet or an email chain. Overhead allocation “figured out” at year-end with no pre-agreed method. Quality metrics defined, or redefined, unilaterally by the managing partner. No partner right to audit the inputs. And any formula whose most senior beneficiary is also its only fluent interpreter.
A partnership cannot be built on a compensation formula that one partner alone can explain. Dr. Taylor never took the recruiter’s dinner. The practice’s counter-offer arrived first, and it contained something better than money: arithmetic she could check.
Compensation per work relative value unit runs $35 to $50 for general pediatrics. Compensation-to-collections ratios of 25 to 35 percent mark the conventional healthy band, with 20 to 40 percent the boundary of defensible. Production benchmarks put the national median near 5,600 units per full-time equivalent, with a 10th-to-90th range of 3,200 to 8,800.
The methodology, and the vote required to change it, belong in the partnership agreement at the supermajority tier, so no future majority can reprice a colleague's economics on a simple vote. Each partner's individual numbers belong in an employment agreement. Any formula a smart physician cannot recompute from the monthly package within 30 minutes is a resignation letter in escrow.
Affordability is a separate gate from pricing. The governing inequality admits no exceptions: overhead percent plus compensation percent cannot exceed 100 percent of collections. Break-even revenue equals total loaded provider cost divided by one minus the overhead rate, and the productivity threshold is break-even raised by whatever margin the practice decides to keep.
The Partnership Agreement Analyzer scores an existing agreement against the framework this series is built on — or schedule a discovery call to work through it with PMI. The full framework, with the arithmetic, lives in the textbook Pediatric Practice Management: The Fundamentals.