State Voting Patterns in the 2016 Presidential Election and Uninsured Rates in Non-elderly Adults.

INTRODUCTION

Medicaid expansion and the health insurance “Marketplaces” of the Affordable Care Act (ACA) reduced the number of uninsured individuals to record lows in 2016.[1] Uninsured rates may have since increased because of recent federal policies, including nullifying the individual mandate and reducing federal funding for Marketplace advertising and navigator programs.[2] Even the expressed desires of congress and the President to repeal the ACA may have increased uninsured rates through a phenomenon known as “presidential cueing”[3] by decreasing public support for the ACA, particularly in states with a plurality of Republican voters.

State-level policies may have enhanced or attenuated the impact of these federal actions on uninsured rates, with specific policies generally varying based on whether states have a plurality of Republican or Democratic voters.[2, 4] For example, some states have chosen to not expand Medicaid,[2] and some states that initially expanded Medicaid have since applied for “Section 1115” waivers, which would allow for modifications to the Medicaid program such as the implementation of work requirements, that would effectively reduce enrollment.[5] The degree to which states have devoted resources to Marketplace outreach and advertisement strategies to compensate for reduced federal funding also varies substantially along party lines.[2, 4]

Given the varying state responses to federal actions weakening the ACA, we conducted a descriptive study examining recent uninsured rates among adults aged 18–64, stratified by their states’ voting patterns in the 2016 presidential election.

METHODS

We analyzed 2009–2018 data from the U.S. Census Bureau’s Small Area Health Insurance Estimates,[1] stratifying individuals based on their states’ voting patterns. “Blue” (N = 21, including Washington, DC) and “Red” (N = 24) states had a plurality of Democratic or Republican voters in both the 2012 and 2016 presidential elections, respectively. “Purple” states (N = 6) had a plurality of Democratic voters in 2012 and Republican voters in 2016; no Red states “flipped” to Blue in 2016. We compared the number and percent of uninsured individuals living in Blue, Red, and Purple states before and after the 2016 election.

RESULTS

From 2014 to 2016 (compared with 2013), uninsured adults aged 18–64 decreased by 15,803,039 (− 8.3 percentage points [ppts]) overall (Fig. 1). Blue states saw a larger relative decrease in uninsured individuals during that period (− 9.1 ppts; 7,658,286 individuals) compared with those living in Purple (− 8.0 ppts; 2,981,026 individuals) or Red states (− 7.7 ppts; 5,163,727 individuals).

Figure 1

Uninsured rates among adults 18–64 years, stratified by state voting patterns.

From 2017 to 2018 (compared with 2016), uninsured individuals increased by 853,474 (+ 0.3 ppts) overall. While Blue states saw a negligible 8,383 (− 0.0 ppts) decrease in uninsured individuals, Purple and Red states saw increases of 241,475 (+ 0.5 ppts) and 620,382 (+ 0.7 ppts), respectively (Table 1).

Table 1

Number and Proportion of Uninsured Individuals in Blue, Purple and Red States

	2009	2010	2011	2012	2013	2014	2015	2016	2017	2018
All Adults 19–64 years
N	187,746,677	189,379,289	190,888,983	191,640,968	192,461,143	193,600,545	194,584,952	194,808,251	195,788,599	195,883,847
No. Unins. ± MOE	38,791,372 ± 112,495	40,754,555 ± 118,992	40,297,670 ± 125,395	39,791,596 ± 114,163	39,336,439 ± 118,819	31,701,534 ± 107,144	25,700,940 ± 100,278	23,533,400 ± 98,242	23,979,301 ± 101,165	24,386,874 ± 103,467
% Unins. ± MOE	20.7 ± 0.06	21.5 ± 0.06	21.1 ± 0.07	20.8 ± 0.06	20.4 ± 0.06	16.4 ± 0.06	13.2 ± 0.05	12.1 ± 0.05	12.2 ± 0.05	12.4 ± 0.05
∆ No. Unis.	-	+ 1,963,183	− 456,885	− 506,074	− 455,157	− 7,634,905	− 6,000,594	− 2,167,540	+ 445,901	+ 407,573
Blue states
N	83,497,762	84,131,825	84,748,155	85,142,709	85,415,928	85,942,425	86,300,761	86,216,102	86,584,168	86,173,697
No. Unins. ± MOE	15,607,205 ± 72,563	16,417,439 ± 81,121	16,099,414 ± 84,327	15,907,465 ± 76,812	15,704,129 ± 76,454	11,922,166 ± 66,884	9,071,854 ± 61,405	8,045,843 ± 58,291	8,050,627 ± 59,081	8,037,460 ± 59,798
% Unins. ± MOE	18.7 ± 0.09	19.5 ± 0.10	19.0 ± 0.10	18.7 ± 0.09	18.4 ± 0.09	13.9 ± 0.08	10.5 ± 0.07	9.3 ± 0.07	9.3 ± 0.07	9.3 ± 0.07
∆ No. Unis.	-	+ 810,234	− 318,025	− 191,949	− 203,336	− 3,781,963	− 2,850,312	− 1,026,011	+ 4784	− 13,167
Purple states
N	37,011,754	37,232,965	37,474,048	37,506,817	37,590,410	37,721,091	37,849,231	37,876,400	38,042,730	38,128,489
No. Unins. ± MOE	7,132,486 ± 48,112	7,492,888 ± 49,689	7,343,006 ± 52,855	7,207,706 ± 47,914	7,142,070 ± 50,925	5,803,021 ± 45,888	4,542,860 ± 42,748	4,161,044 ± 41,556	4,265,158 ± 43,088	4,402,519 ± 44,129
% Unins. ± MOE	19.3 ± 0.13	20.1 ± 0.13	19.6 ± 0.14	19.2 ± 0.13	19.0 ± 0.14	15.4 ± 0.12	12.0 ± 0.11	11.0 ± 0.11	11.2 ± 0.11	11.5 ± 0.12
∆ No. Unis.	-	+ 360,402	− 149,882	− 135,300	− 65,636	− 1,339,049	− 1,260,161	− 381,816	+ 104,114	+ 137,361
Red states
N	67,237,161	68,014,499	68,666,780	68,991,442	69,454,805	69,937,029	70,434,960	70,715,749	71,161,701	71,581,661
No. Unins. ± MOE	16,051,681 ± 71,238	16,844,228 ± 71,480	16,855,250 ± 76,283	16,676,425 ± 69,551	16,490,240 ± 75,361	13,976,347 ± 70,004	12,086,226 ± 66,766	11,326,513 ± 67,281	11,663,516 ± 69,909	11,946,895 ± 71,987
% Unins. ± MOE	23.9 ± 0.11	24.8 ± 0.11	24.5 ± 0.11	24.2 ± 0.10	23.7 ± 0.11	20.0 ± 0.10	17.2 ± 0.09	16.0 ± 0.10	16.4 ± 0.10	16.7 ± 0.10
∆ No. Unis.	-	+ 792,547	+ 11,022	− 178,825	− 186,185	− 2,513,893	− 1,890,121	− 759,713	+ 337,003	+ 283,379

N Number of adults 19–64 years

No. Unins. ± MOE Number of uninsured individuals ± margin of error, calculated at the U.S. Census Bureau’s customary 90% confidence level[1]

% Unins. ± MOE Percent uninsured individuals ± margin of error, calculated at the U.S. Census Bureau’s customary 90% confidence level[1]

∆ No. Unis Year-over-year change in the number of uninsured individuals

DISCUSSION

From 2017 to 2018, Red states experienced the largest increases in uninsured individuals followed by Purple states, while the numbers were stable for Blue states. The overall increase in uninsured individuals is noteworthy because it reverses a steady years-long trend.

This descriptive study precludes causal inferences, and there are important drivers of insurance rates we did not account for, such as employment rates and the economic and political realities at the local level. Nonetheless, the observed differential changes in uninsured individuals in Blue, Purple, and Red states are noteworthy, and may be partially explained by several factors. Most states that did not expand Medicaid were Red states.[2] This is significant because, in non-expansion states, the Marketplaces represent one of the most important sources of insurance for those seeking coverage, especially low-income individuals. However, Marketplace coverage may be unaffordable to this segment given a lack of Marketplace subsidies for certain individuals below the poverty line (known as the Marketplace coverage gap[6]). Additionally, shortened open enrollment periods and the elimination of cost-sharing subsidies directly affect Marketplace plans.[2]

Substantial modifications to some states’ Medicaid programs are being contemplated. Several Medicaid expansion states have applied for (and have been granted) Section 1115 waivers, with the majority of the waivers that would have the effect of decreasing Medicaid enrollment originating from Red states.[5] Due to various legal challenges (and more recently COVID-19), these waivers have not yet been fully implemented.[5]

Tracking the uninsured rate is vitally important as the policy landscape continues to evolve, and millions of people lose their jobs (and thus their health insurance) because of the COVID-19 pandemic. In particular, actions that further weaken the insurance Marketplaces and Medicaid enrollment could exacerbate already stark regional differences in insurance coverage.