Compensation is the most sensitive conversation in any physician partnership. It touches every nerve: fairness, self-worth, trust, work ethic, and money. Most practices avoid the conversation until the tension is unbearable, then try to fix it under pressure, which almost never ends well.
This article provides a structured framework for evaluating your current compensation model, understanding the alternatives, and implementing a change if one is needed. It is based on work we have done with dozens of physician groups across multiple specialties, and it reflects what actually works in practice, not just in theory.
The Equal-Split Model
In an equal-split model, net practice income is divided equally among all partners regardless of individual production. If the practice earns $1.2 million in distributable income and there are four partners, each partner receives $300,000.
When it works
Equal splits work well under a specific set of conditions. First, the group is small, typically two to four partners. Second, all partners have roughly similar production levels, within 10-15% of each other. Third, the partners share administrative duties relatively equally. Fourth, there is high mutual trust and a shared sense of purpose. Fifth, the practice is in a specialty where individual production variance is naturally low (e.g., hospitalist groups with standardized shift structures).
The appeal of the equal split is simplicity. There is nothing to calculate, nothing to argue about, and no complex formula to explain to a new partner. It sends a message that the group values collaboration over individual performance. In practices where it works, it creates a cohesive culture and reduces internal competition.
When it breaks down
Equal splits fail when production variance grows. Here is the math that creates resentment. Imagine a three-partner primary care practice with $900,000 in distributable income. Under equal split, each partner takes $300,000.
Partner A generates 6,500 wRVUs per year (90th percentile for family medicine, per MGMA). Partner B generates 5,200 wRVUs (60th percentile). Partner C generates 4,000 wRVUs (25th percentile). Partner C works fewer hours, sees fewer patients, and generates significantly less revenue, but takes home the same amount.
Partner A is subsidizing Partner C by roughly $75,000-$100,000 per year (the difference between what each contributes to revenue versus what each receives). Over five years, that is $375,000-$500,000 in foregone compensation for Partner A. At some point, Partner A starts looking at employment opportunities where their production is rewarded. And when Partner A leaves, the remaining partners suddenly have to absorb their patient panel, their call coverage, and their revenue, while their overhead stays the same.
This is the most common death spiral for equal-split practices: the highest producer leaves, the economics worsen for the remaining partners, the next highest producer leaves, and eventually the practice collapses or merges.
The warning signs
If any of the following are true, your equal-split model may be under strain. Partners are openly discussing production differences. One or more partners have reduced their hours or patient volume without group consensus. The production variance between the highest and lowest producer exceeds 25%. You are having difficulty recruiting new partners because the comp model is unattractive to high producers. Any partner has explored outside opportunities.
The Production-Based Model
In a pure production-based model, sometimes called 'eat what you kill,' each provider's compensation is directly tied to their individual production, usually measured by collections or wRVUs.
Common structures
Percentage of collections: Each provider receives a fixed percentage (typically 40-55% for primary care, 35-50% for surgical specialties) of their personal net collections. This model is straightforward and directly ties compensation to revenue generation.
wRVU-based: Each provider is compensated based on their wRVUs multiplied by a dollar-per-wRVU rate. MGMA publishes median compensation-per-wRVU data by specialty. For 2025, the median for family medicine is approximately $48 per wRVU; for orthopedic surgery, approximately $62; for cardiology, approximately $68. The practice sets its rate based on what it can afford after covering overhead.
Collections minus overhead: Each provider is responsible for their direct overhead (MA salary, malpractice, supplies) plus a share of indirect overhead, and takes home whatever is left. This model gives providers maximum transparency into their economics but requires robust overhead allocation methodology.
When it works
Production-based models excel in groups with significant production variance, practices that want to recruit high-volume providers, specialties where individual production is easily measured, and practices that value accountability and incentive alignment.
These models are popular in surgical specialties, procedural cardiology, gastroenterology, and other fields where individual production varies widely based on referral base, surgical skill, and work ethic.
When it breaks down
Pure production models create their own problems. First, they discourage teamwork. If every patient seen by a colleague is a patient not seen by you, there is a disincentive to share referrals, cross-cover, or collaborate on complex cases. Second, they undervalue administrative work. The partner who spends 10 hours per week on practice management, credentialing, quality improvement, and strategic planning generates zero wRVUs for that time. Over a career, the 'administrative partner' subsidizes the group with unpaid management labor while earning less. Third, they can incentivize volume over quality. When compensation is purely tied to production, there is a financial incentive to see more patients more quickly, which can conflict with patient care goals and quality metrics. Fourth, they ignore the factors discussed in our article on provider profitability: payer mix, overhead consumption, and ancillary revenue. A provider with high wRVUs but a poor payer mix may generate less revenue per wRVU than a lower-volume colleague.
The Hybrid Model
Most successful multi-provider practices land on some form of hybrid model that combines elements of both approaches. The specifics vary, but the structure typically includes three components.
Component 1: Base salary (60-70% of total compensation)
A guaranteed base salary set at or near the MGMA median for the provider's specialty and experience level. This provides income stability, makes the practice attractive to recruits, and reflects the minimum value of the provider's clinical, administrative, and on-call contributions.
How to set it: Use MGMA data filtered by specialty, geographic region, and practice type (single vs. multi-specialty, academic vs. private). Adjust for local cost of living. A family medicine physician in San Francisco and one in rural Tennessee should not have the same base, even if their wRVU production is identical. MGMA's 2025 data shows median total compensation for family medicine at approximately $280,000 nationally, but the range spans from $230,000 (25th percentile) to $340,000 (75th percentile), and geographic adjustment factors can shift those numbers by 10-25%.
Component 2: Production bonus (20-30% of total compensation)
A bonus pool tied to individual production above a defined threshold. The threshold is typically set at the wRVU level implied by the base salary. If you are paying a base salary at the 50th percentile, the production threshold is set at the 50th percentile wRVU benchmark. Every wRVU above that threshold earns a bonus at a specified rate.
Example: Base salary of $280,000 for a family medicine physician, tied to a threshold of 4,800 wRVUs (approximately MGMA median). The bonus rate is $45 per wRVU above 4,800. If the physician produces 5,800 wRVUs, the bonus is 1,000 x $45 = $45,000. Total compensation: $325,000.
This structure rewards high producers without penalizing those who produce at the median. It also creates a natural governor: the bonus is capped by the provider's ability to see patients, and the per-wRVU rate can be set to ensure the practice can afford the payout after covering overhead.
Component 3: Quality, citizenship, and leadership adjustments (5-15% of total compensation)
This is the component that accounts for everything production metrics miss. It typically includes quality measures (patient satisfaction scores, clinical quality metrics, documentation timeliness), citizenship behaviors (mentoring, committee participation, covering for colleagues, participating in practice development), and leadership responsibilities (managing partner duties, service line leadership, compliance oversight).
These adjustments can be structured as a bonus pool, a point-based system, or a discretionary allocation by a compensation committee. The key is that the criteria are defined in advance, transparent, and applied consistently.
Example: A $20,000 annual citizenship pool divided among partners based on a point system. Points are awarded for serving on committees (5 points each), mentoring new providers (10 points), serving as managing partner (20 points), and achieving quality targets (5-15 points). Points are tallied annually, and the pool is distributed proportionally.
Benchmarking Against MGMA
MGMA (Medical Group Management Association) publishes the most widely used compensation benchmarks in healthcare. Their annual DataDive reports cover compensation by specialty, region, practice type, and academic versus private setting.
When using MGMA data, keep several things in mind.
Median is not the target. The median represents the 50th percentile. Half of physicians earn more, half earn less. Where your compensation should fall depends on your geography, payer mix, patient volume, and practice profitability. A practice with a strong commercial payer mix and low overhead can afford to pay above median. A practice with heavy Medicare/Medicaid exposure and thin margins may need to target the 25th-40th percentile.
Total compensation includes more than salary. MGMA's total compensation figures include base salary, productivity bonuses, quality bonuses, on-call pay, and employer retirement contributions. Compare apples to apples. If your base salary is at the 50th percentile but you offer no retirement match or bonus opportunity, your total compensation is probably below the 25th percentile.
Regional adjustments matter. MGMA provides regional breakdowns (Eastern, Western, Midwest, Southern) and metropolitan versus non-metropolitan data. Use the most relevant subset. National data can be misleading for a practice in a high-cost or low-cost market.
How to Structure the Partner Conversation
Changing a compensation model is not a financial exercise. It is a change management exercise. The math is the easy part. Getting buy-in is the hard part. Here is the approach that works.
Step 1: Present the data, not a proposal
Before proposing any changes, share the objective data with all partners. This includes current production by provider (wRVUs, collections, encounters), current compensation by provider, MGMA benchmarks for your specialty and region, overhead allocation by provider, and the provider profitability analysis. Let the partners sit with the data for at least two weeks before any discussion of changes. The data should speak for itself. If the current model is fair, the data will show it. If it is not, the data will show that too.
Step 2: Agree on principles before discussing mechanics
Before jumping to formulas, get alignment on the principles that should guide compensation. Ask questions like: Should compensation be correlated with individual production? If so, how strongly? How should we value administrative and leadership contributions? Should we benchmark against MGMA? If so, which percentile do we target? How much income stability do we want versus how much performance upside?
These conversations are less contentious than formula discussions because they are abstract. But the principles you agree on will directly inform the model you build.
Step 3: Model multiple scenarios
Build three or four compensation models and run the actual numbers for each partner under each model. Show each partner what they would have earned last year under each scenario. This removes the guesswork and lets partners evaluate concrete outcomes rather than theoretical frameworks.
Step 4: Phase the transition
Do not flip the switch overnight. If you are moving from equal split to a hybrid model, phase it over two years. Year 1: 75% equal split, 25% production-based. Year 2: 50/50. Year 3: full implementation of the new model. This gives partners time to adjust their production patterns and reduces the shock of a sudden compensation change.
Implementation: The First 12 Months
Once the group agrees on a model, here is the implementation roadmap.
Month 1-2: Document the model. Write a compensation plan document that specifies every component, formula, threshold, and adjustment. Have it reviewed by a healthcare attorney (Stark Law and Anti-Kickback considerations apply to physician compensation). Have each partner sign it.
Month 3: Set up the reporting. Ensure your practice management system can produce the production and collections reports needed to calculate compensation under the new model. If it cannot, you need to either reconfigure the system or build supplemental reporting. Test the reports against known data to verify accuracy.
Month 4-6: Run in parallel. For the first quarter, calculate compensation under both the old model and the new model. Share both calculations with all partners. This builds confidence in the new system before it affects anyone's paycheck.
Month 7-12: Go live with the phase-in. Begin paying under the new model at the agreed phase-in percentage. Track production and compensation monthly. Hold quarterly reviews where all partners see the full data.
Annual review. At the end of each fiscal year, review the model's outcomes against the practice's financial goals and partner satisfaction. Adjust thresholds, bonus rates, or pool allocations as needed. No compensation model should be set in stone; the practice that was right for this model two years ago may have different needs today.
A Real-World Transition
A five-physician orthopedic surgery practice came to us with an equal-split model they had used for 12 years. Two surgeons were high-volume joint replacement specialists generating 7,500+ wRVUs each. Two were sports medicine physicians generating around 5,000 wRVUs. One was a hand surgeon generating 4,200 wRVUs. The two joint replacement surgeons were openly frustrated and one had begun interviewing with a hospital-employed group.
We ran the full analysis: production, collections, payer mix, overhead allocation, and ancillary revenue (the joint replacement surgeons generated significant implant and facility fee revenue for the practice's ASC). The data showed that the two highest producers were each contributing approximately $350,000 more in net revenue than the lowest producer, yet all five earned the same amount.
We designed a hybrid model with a base salary at MGMA 40th percentile for each subspecialty, a production bonus above a threshold set at the 50th percentile wRVU benchmark, ASC profit sharing allocated by case volume, and a $100,000 annual citizenship pool allocated by committee.
We modeled the outcomes for each surgeon under the new model. The two joint replacement surgeons would earn approximately 25% more. The two sports medicine physicians would earn roughly the same (their production was near the median, and the base salary provided stability). The hand surgeon would earn approximately 10% less, partially offset by citizenship credits for the administrative work he contributed.
The transition was phased over 18 months. In Year 1, 60% of compensation followed the old model and 40% followed the new model. In Year 2, the new model was fully implemented. The joint replacement surgeons stayed with the practice. The hand surgeon increased his clinical hours modestly. And the sports medicine physicians, who had worried about the change, found that their total compensation was essentially unchanged because the base salary protected their income floor.
The practice retained its highest-revenue producers, resolved years of simmering tension, and established a framework that could accommodate future partners without renegotiating the entire structure.
Final Thought
There is no universally correct compensation model. The right model for your practice depends on your size, specialty, culture, financial position, and strategic goals. What is universal is the need to evaluate the question honestly, with real data, and to revisit it regularly. The practices that fail are the ones that avoid the conversation until it is too late. The practices that thrive are the ones that treat compensation design as an ongoing discipline, not a one-time decision.