Conceptualizations of “public health” and government action among U.S. adults in 2025

Abstract

Introduction

Understanding how Americans think about public health and the role of government institutions is crucial for informing effective health communication in a polarized political climate.

Methods

We deployed an open-ended national survey of 1004 adults in October 2025 to assess public perceptions of these topics.

Results

Respondents associated the term “public health” with health for everyone (44%) and health of everyone (23%), alongside healthcare (20%) and other themes. The most common perceived government benefit was facilitating access to healthcare (39%)—particularly by addressing cost (16%). Among perceived government harms, more Republicans than Democrats cited overregulation (16% vs 8%, P < 0.01), limiting personal freedoms (16% vs 6%, P < 0.001), and inefficiency (10% vs 4%, P < 0.01); more Democrats than Republicans mentioned dismissing science (7% vs 2%, P < 0.01) and defunding (7% vs 2%, P = 0.02). Weighted analyses yielded minimal substantive differences from unweighted findings.

Conclusion

Findings reveal both cross-partisan common ground in term associations and divergent concerns regarding government action. Public health messaging should leverage shared priorities and acknowledge partisan concerns about government roles.

Key points.

• American adults responding to an open-ended survey most often associated “public health” with themes of health for everyone (44%), health of the public (23%), and healthcare system components (20%).

• There were few significant partisan differences about potential benefits of government intervention, but multiple significant differences about potential harm.

• Public health communication could leverage healthcare access as an entry point to bring people into a broader supportive dialogue about public health and its role.

Introduction

Political action related to public health shapes and is shaped by public understanding, opinions, and dialogue.¹ While the term “public health” can accurately signify a range of definitions, the effectiveness of public dialogue that leads with this term may depend on people's immediate conceptual and emotive associations with public health. The range of public conceptualizations of public health has been known to hinder effective dialogue,² exacerbated by different sectors' use of divergent rhetoric for distinct purposes.³ Furthermore, people increasingly engage in public health discourse through rapid, abbreviated, and politically polarized communication⁴ that may elicit automatic cognitive or affective associations rather than deliberative processing.⁵

The field of public health has a long history of defining, refining, and debating its foundational concept. Winslow's 1920 definition of public health as “the science and art of preventing disease, prolonging life, and promoting health through the organized efforts and informed choices of society, organizations, public and private communities, and individuals”⁶ is among the most often cited. In 1988, the Institute of Medicine offered an expanded definition that also encompassed social determinants of health, highlighting the importance of collective action to “assure the conditions in which people can be healthy.”⁷ These communal values have been described as America's “second language,”⁸ which the public health field has embraced but which the American public may find less natural than its “first language” of individualism.⁹ More recently, the American Public Health Association offered another definition emphasizing the protection and promotion of individual and community health¹⁰—a concept whose breadth still allows for ongoing dialogue.

Empirical research has identified notable gaps between how U.S. laypeople and field professionals understand the term “public health.”¹¹ In a national, open-ended 1996 survey about the meaning of “public health,” fewer than 4% of responses aligned with how researchers defined the field.^11,12 A 1999 national survey drew similar conclusions; over half of respondents “could not define public health as either protecting the population from disease or policies and programs that promote healthy living conditions for everyone.”¹³ However, little subsequent research has attended to the consequences of these conceptual gaps, how they may manifest in immediate associations with the term, or their implications among people with divergent political views.

As the sociopolitical context of public health research and practice has changed substantially with COVID-19 and its aftermath,¹⁴ rapid policy changes,¹⁵ increasing polarization,¹⁶ rising institutional mistrust,¹⁷ and quick-consumption media,¹⁸ there is a need to examine dominant associations with “public health” and opinions on government action among the lay public. While the COVID-19 pandemic increased the visibility of public health, this timeframe was also characterized by intensifying ideological divisions about government authority, individual liberty, and scientific expertise.^19,20 Early 2025 saw major leadership and agenda shifts within federal health agencies.¹⁵ Public trust in federal public health institutions showed sharp partisan divides in a March 2025 survey,²¹ while other recent polls identified substantial partisan alignment on priority issues such as “improving healthcare access and affordability” and “preventing chronic diseases.”^21,22 These disparate findings suggest a nuanced reality of Americans' current views of government institutions and their involvement in public health.

Understanding this nuance is crucial for informing effective public health communication in the current political climate, but existing evidence may not sufficiently characterize these views. Prior surveys are limited by discrete response categories that fail to capture the range of conceptualizations the public might hold. Furthermore, while public opinion surveys often capture priorities for government action, they infrequently address potential harms—a particular issue of concern in a politically polarized climate. Additionally, such surveys rarely engage open-ended “word association” approaches, used in linguistic and psychological studies to understand how a broad range of concepts relate in people's mental lexicons.²³ To address these gaps, we conducted a nationally representative open-ended survey to 1) explore Americans' conceptual associations with the term “public health” and 2) understand how they characterized government actions in public health in late 2025. These inquiries can provide an important foundation for communicating about public health, dialoguing across political parties, and shaping future public responses to political action under the banner of “public health.”

Data and methods

We conducted an exploratory mixed-methods analysis²⁴ of a cross-sectional national survey deployed in October 2025.

Participants and recruitment

We recruited participants through Dynata, a frequent facilitator of online recruitment for peer-reviewed survey research that uses non-probability quota-based sampling to model nationally representative samples.^25-27 We aimed to sample 1000 respondents through Dynata's multiple recruitment channels to reflect national distributions of key demographic characteristics; any U.S. resident over age 18 was considered eligible. Dynata compensated all participants according to channel-specific incentive structures. Survey participation was anonymous, and all methods were approved by the Children's Mercy Kansas City Institutional Review Board (#3748).

Survey design

We elicited participants' top-of-mind associations with the term “public health” and their opinions about government involvement in public health through four open-ended questions:

• (Q1) What is the first thing that comes to mind when you hear the term “public health″?

• (Q2) [Optional]: What else comes to mind about the term “public health″?

• (Q3) In what ways can government action in public health be beneficial?

• (Q4) In what ways can government action in public health be harmful?

Participants also self-reported demographic characteristics. All questions were asked in English to maximize comparability of elicited concepts, although we accepted responses in any language.

Qualitative analysis

We qualitatively coded open-ended responses using Dedoose 10²⁸ before viewing demographic data. We developed descriptive inductive codes for Q1 and Q2 responses, aiming to capture all distinct ideas present in the data. We created separate inductive codebooks to thematically characterize Q3 and Q4 responses, applying a gerund strategy to define action-oriented codes (beginning with “-ing” verbs).²⁹ EH and LY created codes, meeting regularly to iteratively merge and classify codes into hierarchical thematic categories (“parent codes”) while refining codebook definitions and application rules²⁹ and maintaining an “audit trail”³⁰ of process notes and previous codebook iterations. We conducted simultaneous coding, applying all relevant codes to a given excerpt, such that no codes were mutually exclusive. We also conducted constant comparisons of qualitative data within and between demographic groups following quantitative analyses to explore and explain statistically significant differences.

Following full team consensus on the hierarchical codebook, EH reviewed all Q1/Q2 responses and code applications to ensure codes were aligned with the data; LY and EH reviewed Q3 and Q4 code applications for a selection of excerpts through second-round coding, discussing minor disagreements until reaching consensus.³⁰ To ensure dependability of code applications, EC was oriented to codebook structure, code definitions, and application rules, and independently reviewed a random 20% sample of excerpts using each of the 3 codebooks (Q1/Q2, Q3, and Q4). EC also conducted a critical review of the codebooks and code applications to assess confirmability, checking whether codes encompassed an appropriate range of concepts and meaningfully reflected participants' statements. Subsequently, a study team member not involved in coding (PW) considered our audit trail updates, key analytic decisions, and overall interpretations to ensure conclusions were supported by the raw data.

If the meaning or relevance of a response was initially unclear, the coder examined the participant's entire response corpus to screen the response quality in the context of their answers to all four questions. We coded responses “unclear/unrelated” if they were clearly not good faith answers (eg, random letters or responses of “good” to all four questions), and we only coded content pertinent to the question asked.

Quantitative analysis

We tabulated the total number of participants engaging each code for each question; we assessed demographic variation across the 10 most frequent Q1/Q2 parent codes and conceptually important subcodes, treating Q1/Q2 combined content as a single response. We also assessed differences by political affiliation for Q3 and Q4, focusing on topics discussed by n≥45 participants. All compartive analyses used $chi-squared$ tests (Fisher's exact tests for cells with <5 observations) using Cramer's V with bootstrapped 95% confidence intervals as a measure of effect size. For overall tests yielding P < 0.05, we additionally performed post-hoc homogeneity of proportions tests using Holm's P-value adjustment for multiple comparisons.³¹

We assessed sample representativeness relative to U.S. population benchmarks calculated from American Community Survey (ACS) 2024 microdata accessed via IPUMS USA,³² restricted to adults aged 18 and older and weighted using IPUMS' provided person weights. Post-stratification weighting procedure details are described in the Supplemental Methods section of the Appendix. We conducted all quantitative analyses using R.³³

Results

Self-reported demographic characteristics are reported in Appendix Table A1. The sample was broadly representative of the U.S. adult population, with statistically significant differences addressed through post-stratification weighting (see Appendix Table A2).

Salient associations with the term “public health” (Q1/Q2)

Our Q1/Q2 codebook identified 12 parent codes representing respondents' top-of-mind thematic associations with the term “public health.” Some themes comprised definitions or descriptions of public health (eg, health for everyone; health of the public) or lexical associations (eg, vaccines/COVID/pandemics, low-income/disadvantaged populations), while others captured affective associations (eg, positive, negative, worry/concern with the state of public health). Figure 1 illustrates the 10 most common general themes captured by our parent codes; the mean number of general themes per participant was 1.75 (SD = 0.86). Table 1 displays definitions and example responses of general themes and conceptually important subthemes.

Figure 1.

The 10 most common conceptual associations with the term “public health” among U.S. adult survey respondents (N = 1004) in 2025. Notes: Displayed themes were derived from inductive descriptive coding of open-ended survey responses. Themes are not mutually exclusive.

Table 1.

Major themes in U.S. survey respondents' conceptual associations with “public health” and views on government action in 2025.

Question	Theme	Definition	Examples
Q1/Q2: What is the first thing that comes to mind when you hear the term “public health’? 10 most frequent general themes (parent codes) and examples	Health for everyone	Describing provision of healthcare and/or services for everyone	“Protecting and improving the health of entire community” “for all people” “making sure everyone has the best care”
	Health of the public	Describing the state of community health	“The overall health and wellbeing of everyone” “health of the people” “community well-being”
	Healthcare system components	Referencing facilities, personnel, medication, equipment	“Hospitals and doctors” “A doctor's office” “The County Nurse who is available to all who need health advice” “Overall healthcare system”
	Government structures/actions	Referencing government offices, agencies, policies, and/or regulatory actions	“The CDC and the government working together” “government” “I think about the government's role to protect the public health like creating the FDA and hospitals” “Government is threatening access to it”
	Vaccines/COVID/pandemics	Referencing COVID, pandemics, and/or related public health measures	“Vaccination programs, health education, emergency preparedness, and ensuring access to healthcare for underserved populations” “the whole COVID lockdown” “I think of things like COVID, CDC and public safety measures. It seems like it is the organization that makes people do quarantine. It would all fall into that same stratosphere to me overall”
	Other associations	Concepts not captured by other themes	“the American people” “Past and current leader” “I need to listen closely, can/may impact me or my family”
	Low-income/disadvantaged populations	Referencing low-income populations or services intended for this group	“Low income based”
	General positive	Positive descriptors used without clarification	“getting better” “extremely important”
	Worry/concern about public health	Expressing concern with the current state of public health	“not the best here in the USA” “Recently, it's been dread and stress” “underfunded” “America isn’t very healthy”
	General health	Referencing health without connection to community-level concepts	“Health issues”
Q1/Q2: What is the first thing that comes to mind when you hear the term “public health’? Other theoretically significant themes and examples	Health for everyone subtheme: Services for everyone	Describing affordability or accessibility issues	“Free health services” “Emergency help when needed, free vaccinations”
	Health for everyone subtheme: Actions/activities promoting population health	Describing specific public health activities	“Making sure the neighborhood is clean, like trash pickup and clean water” “We need to be told if there is some kind of new vaccine or if/when there is an epidemic of any kind”
	Actions/activities promoting population health subtheme: Improving, promoting, protecting health	Describing general principles of public health efforts to prevent disease and/or promote health at the population level	“protecting and improving health of communities and populations” “Community health initiatives, disease prevention, and protecting the health of populations through government programs and policies.” “Caring for the public and making sure everyone is healthy.”
	General negative	Negative descriptors used without clarification	“corruption” “Indoctrination and fear” “Hive mind” “ANGER”
Q3: In what ways can government action in public health be beneficial? Select theme and examples	Facilitating access to healthcare/services for all	Describing strategies for enabling the general population to access care	“Give more options for care, better care, and doctors” “if the government made health insurance [sic] AFORDABLE [sic] or free WITHOUT losing quick access to anything, then that would be optimal”
Q4: In what ways can government action in public health be harmful? Select themes and examples	Hindering access to healthcare/services/insurance	Describing efforts and/or policies that would interfere with the general population's access to healthcare or related resources	“They could take it away” “preventing homeless or low income people from accessing good health care” “Overregulation, mismanagement, or delayed responses can limit access to care”
	Hindering access to healthcare/services/insurance subtheme: Increasing costs	Describing efforts and/or policies that would increase the costs of healthcare and/or make care unaffordable	“by raising health premiums and removing existing programs” “By taking Insurance away and raising the prices, you have low-income people who cannot afford everything” “Pass a bill that cuts government spending towards healthcare, guts healthcare from millions, and raises premiums for private insurance companies”

Themes most often reflected definitional concepts, such as health for everyone (44.2%), which described what public health is by invoking health-promoting services, policies, or interventions (eg, “Vaccinations, disease prevention, clean water, and health education”). Its subthemes included services for everyone (n = 160, 15.9%), which highlighted affordability and accessibility issues (eg, “Making sure everyone has the best care and not just one person with better money”), and actions/activities promoting population health (n = 239, 23.8%), which cited specific public health activities (eg, “Public health is the practice of protecting and improving the health of populations by promoting or teaching healthy lifestyles, preventing disease and injury, and responding to public health threats through some sort of organized efforts.”). The action-forward quality of health for everyone distinguished it from health of the public (22.8%), which described the health status of the general population (eg, “health concerns of people of the community,” “I think of diseases that can spread among the public,” or “epidemic”). Only 9.1% of respondents (n = 91) specifically mentioned improving/promoting/protecting health, which emerged as the concept most closely aligned with field-standard definitions of public health. These responses ranged from general descriptions such as “preventing disease and prolonging health” to more extensive articulations of the definition of public health (eg, “The first thing that comes to mind is the collective effort to keep communities safe and healthy-through things like disease prevention, clean water, vaccinations, and access to medical care. It's about protecting everyone's well-being, not just treating illness after it happens”).

Many conceptual associations could not be clearly classified as either definitional or lexical themes. Healthcare system components (19.8%) described responses that listed elements of the healthcare system (“Affordable doctor appointments,” “Nurse,” “overwhelmed workers,” or “Public health is when you can go to your local urgent care center for medical health as well as an emergency visit at the [sic] your local hospital”). Conceptual associations with government structures/actions (18.1%) were also common and included a range of single-word responses (“Policy,” “CDC”); explanations of government roles in public health (“I think in terms of public policy with regards to improving health. That could be improving the environment (ie, reducing pollutants), making sure vaccines are readily available. When needed, encouraging the use of masks (ie, pandemics, etc.)”); and affectively charged statements (“Questionable edicts,” “government control,” “Our tax dollars at work helping people”). Vaccines/COVID/pandemics (14.2%) featured in many responses that also ranged from descriptive or technical (“Vaccines for fighting coronavirus,” “covid, the flu and other contagious diseases”) to value-laden (“I am concerned about people who don't believe in vaccines who could end up spreading disease.”). Responses often explicitly invoked the COVID-19 pandemic as a strong conceptual association, with a very small number expressing memorable institutional skepticism (“forced muzzling,” “The plandemic that was used for depopulation…and it worked”).

Some responses contained more distinctively affective content (n = 116, 11.6%), which infrequently co-occurred (n = 3, 0.3%) with field-aligned definitions. Related themes included general positive (5.9%), general negative (1.8%), and worry/concern with public health (4.3%). Many responses coded as general positive evoked a sweeping but nonspecific sense of importance, such as “vital” or “Very important but no longer valued by too many in our population”—indicating some respondents' significant support for high-level concepts of public health. Conversely, general negative sentiments included a range of more distinct concerns, such as: “untrustworthy,” “It's being taken advantage of,” “Trumps’ [sic] lies and delusional take on literally everything,” “General bullshit,” and “Interference in your life.” Worry/concern with public health captured statements in which “public health” as such was clearly not the object of negative sentiment; rather, respondents expressed “concern,” with several stating it is “currently in bad shape,” or “I get mad because I feel like it's not taken care of enough.”

Few Q1/Q2 themes differed significantly by demographic characteristics. We found no significant differences by political affiliation in the proportion of responses flagged as low-quality. Results of significant post-hoc tests for all questions are displayed in Table 2, with detailed frequencies and complete analysis results provided in Appendix Tables A3–A6.

Table 2.

Summary of significant thematic differences in “public health” associations and government perceptions by demographic characteristics.

Question	Theme	Higher-endorsement group(s)	Reference group	V [95% CI]a	P b
Q1/Q2: What is the first thing that comes to mind when you hear the term “public health”?	Health FOR everyone	$100-$150 K earners	<$50 K earners	V = 0.12 [0.07-0.19]	0.02
	Government structures/actions	Adults >75 years old	35-44-year olds	V = 0.14 [0.10-0.22]	0.02
		White respondents	Black respondents	V = 0.13 [0.10-0.19]	0.045
		Bachelor's graduates	High school or less	V = 0.14 [0.10-0.21]	<0.01
		Master's graduates			0.01
	General positive	$100-$150 K earners	<$50 K earners	V = 0.12 [0.08-0.20]	<0.01c
		Bachelor's graduates	High school or less	V = 0.17 [0.13-0.25]	0.03
		Master's graduates			<0.01
		Doctoral graduates			0.05
	Personal health	Democrats	Republicans	V = 0.10 [0.07-0.15]	0.03c
		Independents			0.02c
	Population-level health promoting actions	>$150 K earners $100-$150 K earners	<$50 K earners	V = 0.17 [0.12-0.24]	<0.01 <0.001
		Master's graduates	High school or less	V = 0.19 [0.14-0.26]	<0.001
Q3: In what ways can government action in public health be beneficial?	Addressing public need	Republicans	Democrats	V = 0.11 [0.06-0.18]	<0.01
Q4: In what ways can government action in public health be harmful?	Over-regulating, overreaching, controlling too much	Republicans	Democrats	V = 0.13 [0.08-0.21]	<0.01
	Limiting personal freedom/rights	Republicans	Democrats	V = 0.14 [0.09-0.21]	<0.001
	Acting inefficiently	Republicans	Democrats	V = 0.12 [0.07-0.20]	<0.01
	Dismissing or lacking professional/scientific expertise	Democrats	Republicans	V = 0.11 [0.06-0.17]	<0.01
	Withdrawing/withholding funds	Democrats	Republicans	V = 0.10 [0.06-0.16]	0.02

Abbreviations: CI, confidence interval.

^aCramer's V reports effect size for the overall test.

^bAll P-values are reported from post-hoc pairwise comparisons.

^cNot statistically significant (P≥0.05) in weighted analyses; see Appendix Table A4.

Potential benefits or harms of government action in public health (Q3/Q4)

Figure 2 displays the most common Q3 and Q4 themes by political affiliation. The most commonly mentioned beneficial government action was facilitating access to healthcare/services for all (n = 396, 39.4%), which described specific activities such as “provide coverage for those without insurance. Make sure all school children are immunized for free.” Its most common subtheme was controlling costs/increasing affordability of care (n = 159, 15.8%), which similarly referenced ways the government could address high costs of care and insurance coverage (“they could make health care free for the working class citizen,” “Help people who can't afford insurance”). A significantly higher proportion of Republicans (46.6%) than Democrats (33.7%; P  < 0.01) mentioned the benefit of addressing public need, for example, by preventing or controlling disease/injury (“Be proactive with research that may prevent health issues. Monitor health insurance companies on how patients are treated. Also do the same for medical institutions” or by keeping the public informed/disseminating information (“communication and tips on what's going around and tips on how to stay healthy, “Monitoring potential health risks and keeping citizens informed with credible information.”).

Figure 2.

Most common perceived benefits and harms of government action in public health reported by U.S. adult survey respondents (N = 1004) in 2025, by political affiliation. Notes: Themes reflect emergent categories from inductive descriptive coding of open-ended responses; only themes endorsed by at least 45 participants are displayed. Themes are not mutually exclusive. Brackets indicate statistically significant pairwise differences by political affiliation (Holm-adjusted P < 0.05; see Methods). Respondents identified more distinct categories of potential government harms than benefits.

The most common perceived government harm was hindering access to healthcare/services/insurance (n = 171, 17.0%), whose most common subtheme was increasing costs (n = 76, 7.6%) (eg, “Allowing health insurance free reign [sic] over what they charge and the ability to deny treatments,” “continuing to overcharge the American people for healthcare,” “Could be ruinously expensive,” “COST ON BOTH MEDICINE AND SERVICE,” “Dictating on the tax payers [sic] decision on having healthcare when most Americans can't afford coverage.”). Higher proportions of Republicans than Democrats mentioned potential government harms of over-regulating/over-reaching/controlling too much (“to [sic] many overreaching policies that raise the cost of healthcare” (15.8% vs 7.7%, P  < 0.01); limiting personal freedom/rights (“Making too many rules and laws that make it impossible for people to decide for themselves and their own bodies.”) (16.4% vs 6.1%, P < 0.001); and acting inefficiently (“They can cause a monopoly, overcharge or overtax, and be inefficient.”) (10.4% vs 3.8%, P < 0.01). Conversely, more Democrats than Republicans mentioned dismissing or lacking professional/scientific expertise (“By not accepting science and mandating harmful policies or spreading harmful ideas or misinformation”) (7.1% vs 1.8%, P < 0.01) and withdrawing/withholding funds (“Pass a bill that cuts government spending towards healthcare, guts healthcare from millions, and raises premiums for private insurance companies”) (6.9% vs 2.1%, P = 0.02).

All raw data, including codebooks, excerpts, and final code applications for Q1/Q2 and Q3/Q4 are provided in the Supplementary Data File.

Weighted analyses

Weighted analyses yielded substantively similar results. Overall code application frequencies differed by <2% between unweighted and weighted samples and effect sizes were consistent across analyses, with Cramer's V differing by no more than 0.05 across all comparisons. Some unweighted pairwise differences were non-significant in the weighted sample (see Table 2). Weighted analyses yielded several additional significant pairwise associations that were consistent in direction with those of unweighted analyses (Appendix Tables A4–A6).

Discussion

In the context of 2025 public health policy shifts and intensifying political discourse, we surveyed a nationally representative sample of 1004 U.S. adults about their conceptual associations with the term “public health” and opinions about government action. Our findings align with previous studies that have suggested a lack of consensus among U.S. laypeople about what public health is;¹¹ they also offer novel insights regarding common conceptual associations, opinions about government action, and implications for engaging the American public.

Less than half of respondents discussed any given theme, suggesting the term “public health” elicits a diverse range of conceptual associations. Only 9% of all respondents—who were likely to have relatively higher income and education—explicitly mentioned themes of prevention and promotion central to APHA's definition of public health.¹⁰ This represents only a 5% increase over the findings of a comparable open-ended survey conducted 30 years ago,^11,12 suggesting that such salient recent events as the COVID-19 pandemic did not substantially move public conceptions of the field toward a field-aligned definition. While promoting field-aligned definitions of public health among the public has been seen as a communication priority in earlier studies,^11-13 our diverse elicited responses suggest that promoting a particular definition may not be as feasible (or as vital) as it has been made out to be, for several reasons.

First, standardizing definitions may not necessarily help the field engage the lay public on already-salient issues of common concern such as healthcare costs and accessibility. Although some survey respondents could have conflated public health with healthcare—which appears to be a long-standing tendency¹¹—healthcare settings are also central for many public health efforts. Rather than focusing on definitional consensus per se, emphasizing public health efforts to promote healthcare access may help Americans appreciate what the field can offer.

Second, our findings suggest that how the public feels is at least as important as what they think about public health. Emotionally expressive associations (11.6%) were about as common as field-aligned public health definitions (9.1%) and provided especially valuable insights about people's most pressing priorities and concerns. Notably, many of people's expressed concerns—about the field's trustworthiness or a sense of interference, for example—are unlikely to be mitigated by simply revising conceptual definitions; in fact, attempting to do so could further entrench negative perceptions of the field's unrelatability and overreach. Before attempting to impose definitions, the field should prioritize engaging with the public's salient feelings and mitigating their underlying concerns.

Finally, lay definitions of public health seem to have little bearing on actual public support for public health activities. A 1996 study noted this phenomenon in its manuscript entitled “‘Public health’: two words few people understand even though almost everyone thinks public health functions are very important.”¹² Respondents of both past surveys and the current study expressed support for specific public health initiatives without necessarily articulating formal population-based frameworks. Meeting the public where they are—learning to speak their “first language”—may be a far more efficient and effective communication strategy than teaching a “second language.” That is, the public health field may need to explicitly highlight the implications of public health activities for individualistic concerns such as personal freedom, privacy, and autonomy, as well as for population-level disease prevention and health promotion.

Our respondents' ideas about government action provide compelling insights about how public health can align its efforts with lay concerns. We identified substantial common ground regarding beneficial government involvement, and the most frequently cited benefits (facilitating access to care and addressing public need) were similar to bipartisan public health priorities identified in a recent survey.²² Notably, nearly half of Republican respondents mentioned the benefits of the government′s addressing public need, while only 5% expressed that the government could do nothing beneficial. Recent surveys have similarly shown that increasing shares of Republicans endorse government responsibility for providing health coverage.³⁴ Indeed, healthcare costs and affordability emerged as common points of concern across the political spectrum in our study, suggesting there could be broad bipartisan support for efforts to facilitate access to care.

Partisan differences were more pronounced when considering harms (though not for the most frequently invoked concern, hindering access to healthcare). These differences provide important insights into risk-based framing as a potential source of political polarization that is rarely addressed in public opinion research of public health. “Negativity bias,” the well-documented tendency to use and attend to negative rather than positive information,³⁵ could explain why the most politically polarized opinions about public health arose when respondents contemplated how the government could cause harm. The public health field should take great care when engaging in topics that could evoke expectations of government harm across party lines, focusing instead on leveraging bipartisan consensus regarding potential benefits.

Our exploratory study has several limitations. While we aimed to achieve representativeness of the U.S. adult population, our online recruitment strategy restricted our ability to reach people with limited internet access. There are inherent limitations to online panel survey procedures, including the difficulty of adequately identifying satisficer responses and (in our recruitment procedure) a lack of data about people who declined to participate or did not complete the survey. As such, we could not assess systematic differences between completers and non-completers; including additional screening measures or attention checks may also have yielded richer responses with a higher degree of confidence in our data quality. Furthermore, response heterogeneity signaled that our data should be interpreted cautiously; respondents' tendencies to answer questions in certain ways could have substantially shaped response content. Thus, the data may not reflect respondents' definitive perceptions of public health beyond immediate associations. Definitive perceptions could be explored with greater confidence in future studies designed to yield richer and more directed responses. Confirming that our sample adequately represented the U.S. political affiliation distribution is particularly challenging, given the range of estimates from existing national surveys such as Gallup³⁶ and Pew.³⁷ Our findings differed significantly from both estimates but represented their approximate midpoints; despite lacking a single reliable benchmark, we believe our estimates are reasonable in light of our weighting procedure using other available demographic information. Responses were also brief, yielding less nuanced information than interviews or focus groups might produce. Furthermore, insights from the manual inductive coding process reflect our study team's interpretations, and other analysts could have interpreted findings differently. However, our hand-coding strategy allowed us to account for response context that could have been overlooked by automated code creation or application methods. Findings that emerged as nonsignificant in the weighted sample should be treated with appropriate caution. Finally, our sample size precluded more detailed investigations of intersectional differences (eg, by education and political affiliation), which could be important to explore in future research.

Conclusion

We should not expect the public to hold a perfect or uniform definition of public health—such a definition does not truly exist. Regardless, the nuances of public health communication are important in an increasingly partisan media environment. Communication strategies should highlight specific salient issues, such as healthcare access and affordability, to mitigate partisan concerns and leverage shared priorities. In this way, the field can “show, not tell” Americans what public health is. Rather than trying to disseminate collectivistic values as a “second language,” the field should make the effort to speak Americans' “first language”^8,9 to advance the common good through public health in a hyper-partisan era.

Supplementary material

Supplementary material is available at Health Affairs Scholar online.

Funding

Research reported in this manuscript was supported by the National Institute of Mental Health (Award Number F31MH136678 to LY), the National Center for Advancing Translational Sciences (Award Number TL1TR002368 to LY), and the National Center for Advancing Translational Sciences (Award Number UL1TR002366 to EC). Funders had no role in conceptualizing the study, collecting or analyzing data, or deciding to publish the manuscript. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.