Less-Efficient ACOs Entering Medicare Shared Savings Program More Likely to Earn Bonuses

Accountable care organizations (ACOs) that entered the Medicare Shared Savings Program (MSSP) with higher, less efficient spending were significantly more likely to earn shared savings bonuses throughout their participation compared with more efficient peers, according to research published in JAMA Network Open.¹ Notably, a 2017 policy change intended to level the playing field did not close that gap, according to the study.

The MSSP, which is the largest Medicare ACO program, includes almost 500 ACOs encompassing 11 million beneficiaries.² Organizations that take part in the voluntary program have the opportunity to share in savings from the program based on beneficiary benchmarks.¹ These benchmarks are the main determinant of success or failure in the program, and they were historically based on the ACO’s historical spending carried forward using national Medicare spending growth—an approach that favored inefficient organizations that had the most spending to cut.

“With historical benchmarking alone, the Shared Savings Program became prone to what is called selection on slopes, where organizations decide to participate according to expected or already achieved reductions in spending,” the authors explained. “As a result, inefficient organizations had greater opportunity for reward vs efficient ones, as the latter can continue to reduce spending only to a point.” A regional benchmark adjustment in 2017 aimed to encourage more efficient ACOs to participate, but whether the adjustment means that efficient organizations can more easily receive bonuses has not been clear, the authors noted.

The new cross-sectional study analyzed 402 ACOs that participated in the MSSP for at least 4 years between January 2013 and December 2020. Researchers developed an ACO spending ratio—comparing observed spending against expected spending after adjusting for patient characteristics—to measure efficiency at program entry and track its relationship with bonus earnings over time.

Among the 402 ACOs studied, the median (IQR) spending ratio was 1.000 (0.993-1.005). Of the ACOs included, 33% earned bonuses in the second agreement year, 167 earned bonuses in the third agreement year, and 183 in the fourth agreement year.

ACOs that were more efficient at entry into the program were significantly less likely to earn bonuses than the least efficient ACOs, even after the 2017 benchmark adjustment. For the most efficient ACOs, defined as those in the lowest quartile of the spending ratio, the likelihood of earning a bonus increased from 24.4% (95% CI, 15.3%-33.4%) before the 2017 benchmark adjustment to 45.2% (95% CI, 35.4%-55.0%) after the adjustment. For the least efficient ACOs in the top quartile of the spending ratio, the probability of earning a bonus was 43.8% (95% CI, 33.7%-53.9%) before the benchmarking change and 60.7% (95% CI, 51.3%-70.1%) after the change.

In the fourth agreement year, 59% of the least efficient ACOs earned a bonus compared with 31.6% of the most efficient—a statistically significant 27.4–percentage point difference.

The financial implications extended beyond simply receiving a bonus. Less efficient ACOs also earned larger bonuses per beneficiary. In the fourth agreement year, ACOs in the lowest efficiency quartile earned a mean of $226 per beneficiary, while those in the highest earned $399. This underscores the structural advantage built into current benchmarking design, according to the authors. An important finding was that while the 2017 benchmarking adjustment did seem to help ACOs regardless of efficiency at entry, it did not eliminate the gap in bonus probability between the more efficient and less efficient ACOs.

The findings emphasize the need for benchmarks that will motivate and reward ACOs across the spectrum of efficiency to take part in the program, the authors explained. One potential solution could be administrative benchmarks, which have an initial base value that is tied to historical spending but then separate out the growth or shrinkage of the benchmark.

“Instead, the benchmark’s growth would reflect factors such as policy goals, broad economic indicators (eg, gross domestic product or consumer price index), and anticipated changes in patient mix,” the authors said. “Policymakers could scale the benchmark’s growth relative to spending to be slower for inefficient ACOs and faster for efficient ACOs, thereby converging benchmarks across ACOs while encouraging broad participation.”

The study was limited by its inclusion only of ACOs that were in the MSSP for at least 4 years, which are ACOs that foresee short-term success in the program. This limits the findings’ generalizability and could introduce homogeneity regarding future bonus earning potential and more conservative estimates of the impacts of efficiency, according to the authors.

“Without aligning incentives for long-term improvement with sustainable financial rewards, the program risks becoming a revolving door for inefficient ACOs that only marginally improve, yet—by virtue of their inflated benchmarks—command substantial bonuses,” the authors wrote. “Akin to what is termed the market for lemons, which describes the unsustainable selection bias seen in insurance markets, preferentially engaging inefficient ACOs may limit the long-term benefits of these voluntary value-based programs.”

References

1. Srivastava A, Shay A, Kaufman SR, et al. Accountable care organization efficiency on entry and shared savings bonuses. JAMA Netw Open. 2026;9(2):e260166. doi:10.1001/jamanetworkopen.2026.0166

2. Medicare Shared Savings Program continues to deliver meaningful savings and high-quality health care. News release. CMS. October 29, 2024. Accessed February 26, 2026. https://www.cms.gov/newsroom/press-releases/medicare-shared-savings-program-continues-deliver-meaningful-savings-and-high-quality-health-care