Disparities in the Impact of High-Deductible Health Plans on Financial Strain

Tuesday, June 25, 2019
Exhibit Hall C (Marriott Wardman Park Hotel)

Presenter: Matthew Eisenberg

Co-Author: Harlan Pittell

Background: Enrollment in high-deductible health plans (HDHPs) – plans with high deductibles and tax-advantaged savings accounts – is increasing rapidly. While the effect of HDHPs on total spending and utilization is well established, less research has focused on the financial strain associated with these plans. On one hand, these plans have been shown to reduce overall healthcare spending which should translate to lower out-of-pocket (OOP) costs to the consumer. On the other hand, these plans include higher levels of cost sharing which could lead to an increase in total OOP spending.

Objective: To estimate the intent-to-treat (ITT) effect of an employer offering a HDHP on health plan spending, out-of-pocket spending, and financial strain.

Data: We use data from OptumLabs (2005-2016) that include detailed individual-level information on enrollment, claims, benefit design, race, education (census block level), and household income. We combine this data with publically available data from the Kaiser Family Foundation (KFF) to simulate the range of effects of HDHPs on total financial strain, after accounting for premiums and employer account contributions. Our sample includes 2,780,346 person-years of data at HDHP offering employers and 12,620,490 person-years of data at employers that never offer a HDHP.

Methods: The study uses an intent-to-treat (ITT), difference-in-differences study design to analyze the effect of an employer offering an HDHP on various spending outcomes. The key identification challenge of such an analysis is the endogeneity of plan selection. Our analysis minimizes bias from selection by focusing on employer offer of an HDHP rather than individual level enrollment. We estimate a series of individual-year-level two-way fixed effects models with employer and calendar year fixed effects in order to identify our main difference-in-differences effect (EmployerOfferxPost). We estimate models on health plan spending, OOP spending, and financial strain (OOP/household income). Additionally, we estimate a series of stratification models examining the effect across race/ethnicities, educational attainment (census block level), and household income. To simulate the total financial strain effects, accounting for premiums and account contributions, we describe the distribution of total financial strain across the joint distribution of median HDHP premiums and account contributions for similarly sized employers in the KFF Employer Health Benefit Survey.

Results: We find that employer offer of a HDHP significantly increases OOP spending (25.3% off a baseline $537) while reducing total (OOP + health plan) spending (5.0% off a baseline $3,930). Additionally, we find that financial strain (OOP/household income) increases from 0.83% to 0.98% after HDHP offer, compared to the control group. We find no significant interaction effects across the SES distribution.

Conclusions: HDHPs significantly increase OOP spending while reducing total utilization and these effects are sizeable. Back of the envelope calculations with our ITT estimates suggest that an individual enrolling in a HDHP can expect to pay $403 more dollars OOP than someone who did not enroll. While premium savings may blunt the increased strain associated with these plans, preliminary analyses of the KFF Employer Health Benefit Survey suggest that premium savings associated with HDHPs are not fully passed on to the workers.