Covered but Still Left behind: Government-Subsidized Coverage without Realized Access and the Path-Dependent Persistence of Unmet Healthcare Needs In South Korea

Background/Purpose: Universal health coverage is an established social right, yet enrollment alone does not guarantee access — a gap with direct implications for economic inequality. The gap is clearest in dual-track systems, where means-tested public programs operate alongside mainstream insurance; the US Medicaid-Medicare parallel structure reflects a similar logic. In South Korea, Medical Aid and the Korea Veterans Health Service (KVIS) provide government-subsidized coverage to low-income and veteran populations outside the National Health Insurance (NHI) framework; NHI covers approximately 97% of the population, yet cost-sharing subsidies have not prevented disproportionate burden among program beneficiaries. Grounded in cumulative disadvantage and path dependency frameworks, this study asked (1) whether unmet healthcare needs exhibit state dependence, and (2) whether socially vulnerable individuals — enrollees in Medical Aid or KVIS — carry persistent excess risk beyond what prior experience alone explains.

Methods: Data came from the Korea Health Panel Survey (2011–2018; n = 10,132 adults). Social vulnerability was defined by enrollment in Medical Aid or KVIS, with NHI enrollees as the reference group. Unmet need subtypes followed Lévesque et al.'s (2013) access framework (approachability, availability, affordability). We estimated dynamic correlated random-effects probit models with initial conditions correction, reporting average marginal effects (AMEs). Subtype analyses applied Benjamini-Hochberg false discovery rate correction (q = 0.05), adjusting for individual-level covariates and survey-year fixed effects.

Results: Unmet need affected 16.3% of the full sample at baseline — but that figure reached 25.8% among socially vulnerable individuals, versus 15.9% for NHI enrollees. Having experienced unmet need once raised the probability of a subsequent episode by 4.16 pp (95% CI: 3.30–5.01), a 25.5% increase relative to baseline — evidence that deprivation does not reset between periods but compounds. Adjusting for this, enrollment in Medical Aid or KVIS showed an additional 3.51 pp penalty (+22.1%): history explains part of the gap, but not all of it. The affordability subtype deepened this contrast. Cost-related barriers affected 16.4% of socially vulnerable individuals versus 5.5% of NHI enrollees (AME = 1.15 pp; p = 0.001) — a threefold disparity that persisted despite subsidized cost-sharing. The approachability subtype told a different story: socially vulnerable individuals actually reported lower rates (1.97% vs. 4.14%; AME = −0.89 pp; p = 0.012), suggesting coverage may reduce navigational barriers even where cost burdens persist.

Conclusions and Implications: A first episode of unmet need is not just a data point — it is a threshold. Once crossed, deprivation tends to compound. Social workers and care navigators — whether in clinics, community organizations, or managed care networks — can intervene before the pattern takes hold. A separate, program-rooted structural penalty persists regardless of prior history. The practitioner’s role shifts from case identification to systemic advocacy: pressing for benefit expansion and confronting provider access constraints that cost-sharing alone has not resolved. For researchers, separating self-reinforcing deprivation from residual structural disadvantage offers a portable diagnostic framework — applicable wherever means-tested programs run alongside mainstream coverage, including the US dually eligible population.