Health Policy$ense

Do High-Deductible Health Plans Bend the Spending Curve?

New study finds slower growth in long-term spending

We know that high deductibles have a significant effect on spending levels, but do they affect spending growth?  In a recent National Bureau of Economic Research (NBER) working paper, LDI Associate Fellow Molly Frean and LDI Senior Fellow Mark Pauly found that spending growth was significantly lower in states where privately insured employees have higher deductibles. The authors analyzed state-specific data on deductibles and various categories of health care spending over 15 years (2002-2016), during which deductibles more than tripled in magnitude and spending growth exceeded 40%.

What's a high deductible?

Generally, a deductible is considered “high” when it exceeds $1,000 for individual coverage or $2,000 for family coverage. The Internal Revenue Service (IRS) has standards for high-deductible plans, which change year to year. Current IRS standards are $1,350 for individual coverage and $2,700 for family coverage.

While the level of spending is important, it is the current growth (or “inflation”) of medical spending that is potentially unsustainable. By giving patients “skin in the game” early in a given year of coverage, deductibles can shape both the care that patients seek and the care that physicians provide. Indeed, multiple studies have shown that spending falls when an individual or a population is exposed to higher cost-sharing in the form of a high or higher deductible. However, Pauly and Frean emphasize that lower spending levels are not the same as lower spending growth.

So, why might we expect deductibles to affect spending growth? For deductibles to lower long-term spending growth, they must play a role in the introduction and use of the new (and costly) technologies that are known to drive spending growth. Consider a patient diagnosed with a condition that can be treated with a low-cost existing standard of care (option A) or a new, more expensive, therapy (option B). Assuming a large price difference between the options, a patient with a high deductible may choose option A after weighing the additional out-of-pocket costs for option B against the value of its additional perceived clinical benefits. The same patient might choose option B when he/she has no deductible or a low deductible, since the incremental cost of option B would be lower, or possibly zero. If many patients in high deductible plans react to new technologies in this way, we would expect spending growth to fall.


The NBER paper, “Does High Cost-Sharing Slow the Long-term Growth Rate of Health Spending? Evidence from the States”, was written by Molly Frean, PhD Candidate, The Wharton School, and Mark Pauly, PhD, Professor of Health Care Management and Business Economics and Public Policy, The Wharton School, University of Pennsylvania.

In their study, Frean and Pauly used national data from the Centers for Medicare and Medicaid Services (CMS) and the Agency for Healthcare Research and Quality (AHRQ). They draw on the state health expenditure accounts released by CMS and the insurance component of AHRQ’s Medical Expenditure Panel Survey (MEPS). The latter includes detailed information on the health benefits offered to employees of about 40,000 public and private sector employers. Across a series of econometric models, they find spending growth is lower in states where (i) more employees are enrolled in private plans with non-zero deductibles; and/or (ii) deductibles are higher in private plans that do have non-zero deductibles. The authors also find that controlling for other measures of cost sharing does not affect the impact of variations in deductibles alone.

It is also worth noting that not all out-of-pocket costs are created equal: like deductibles, copayments and coinsurance also represent forms of consumer cost-sharing that may influence spending patterns. These forms differ, however, with regard to when and how they apply to incurred expenses, making deductibles worthy of study in isolation. 

Next on the authors’ research agenda is to explore this phenomenon in more detail, including looking at specific measures of technology diffusion across health plan types with more granular data. Such work will speak to the actual patient experiences that are underlying the relationships present in the aggregate data.

With health care spending rising faster than GDP, reducing the growth rate is a primary health policy goal. This research suggests that high-deductible health plans may have a part to play in achieving this objective, but more work is needed to understand just how they might operate in doing so. Policymakers who regulate maximum out-of-pocket exposure within health insurance plans and the tax-shielded spending accounts that are often paired with high-deductible plans should consider these and future findings in their decisions. This research is also relevant to other stakeholders in the health economy. Earlier this month, the Kaiser Family Foundation briefed reporters on results from an annual survey showing that the proportion of workers with a general annual deductible reached 85% in 2018, with an average deductible of $1,573 for single coverage. Knowledge of how these rising figures affect long-term spending trends will help employers in making decisions about benefit packages to offer to employees, where the future of high deductible plans is uncertain. Additionally, these findings should be of interest to insurance companies as they design and market different insurance products to firms and to individuals on the ACA exchanges.