Keywords: Cost Growth Targets, Healthcare Expenditure, Policy Evaluation

Background: In 2023, national medical expenditures reached $4.9 trillion, accounting for nearly 18% of the U.S. GDP, growing at rates that exceed economic growth, and placing a significant strain on patients, payers, and American households and businesses.1,2 In response, several states implemented cost-growth target models, setting annual growth caps to limit expenditure growth over the past decade.3 Rigorous empirical examinations of the aggregated effectiveness of these models are critically needed to inform policymaking focused on curbing expenditure growth. To fill this evidence gap, we performed a quasi-experimental evaluation on the impact of cost-growth target models on the growth of U.S. per capita total medical expenditures during 2010-2020.

Objective: This study evaluates the impact of state-level cost-growth target models on total per capita health care spending growth from 2010-2020 using CMS State Health Expenditure Accounts (SHEA) data. 

Methods: We leveraged difference-in-differences methods to evaluate the impact of cost-growth target models using CMS State Healthcare Expenditure Accounts using the most recent data available from 1/1/2010-12/31/2020. Five treatment states with cost-growth policies in varying years were compared to states without these policies (Figure 1). We performed parallel trends testing on pre-policy expenditure trends for each treatment state vs all other control states. We used a fixed-effects regression model to account for unobserved heterogeneity, within-state variability, and time-specific shocks, with all other U.S. states serving as a contemporaneous control group. Models also adjusted for age, sex, race, and ethnicity using U.S. Census data, and robust standard errors were clustered by state.

Results: We included 561 state-year observations, revealing 39.57% mean year-to-year increases in per capita medical expenditure across states from 2010-2020 (p=0.000). Cost-growth target models were associated with a 2.04 relative reduction in total medical expenditure growth (95% CI: -0.69%,-3.40%; p=0.004). Results remained robust in sensitivity analyses excluding 2020 data to account for the impact of COVID-19 on expenditures, where we observed a 1.33 relative reduction in expenditure growth (95%CI: -0.43%,-2.22%; p=0.004). Notably, the reduction was primarily driven by Medicare, with a 2.47 relative reduction in Medicare expenditure growth (95%CI: -0.64%,-4.29%; p=0.009). We cannot definitively conclude the impact of cost-growth policies on Medicaid and commercial spending given non-parallel pre-treatment trends.

Discussion: State cost-growth target policies are associated with small reductions in total medical expenditures, particularly in Medicare. These policies generally rely on weak levers limited to public reporting and voluntary compliance, with only Massachusetts implementing modest financial penalties to health systems. The lack of significant impact on Medicaid and commercial expenditures suggests that these models may need to be paired with additional financial levers tailored to these sectors to achieve broader cost containment goals, while avoiding decrements to quality-of-care by focusing on low-value health services.

Conclusion: State cost-growth target policies were associated with significant reductions in personal healthcare spending. 

Limitations: We used aggregated data, which lacks within-state variations. Short post-policy periods for DE and RI, and heterogeneity in policy designs may not fully capture long-term effects, particularly for payer-specific spending.