Preventing firearm-related deaths and the public good: a contingent valuation study in California

Abstract

Introduction

Firearm-related injuries cause far-reaching harm, yet information about the value the public assigns to the benefits of prevention is limited.

Methods

We surveyed California adults from the Ipsos KnowledgePanel (N = 2870) about their willingness to pay (WTP), in taxes or donations, to prevent firearm homicides, firearm suicides, and deaths from mass shootings. WTP was calculated using a double-bounded dichotomous choice contingent valuation model with a log-logistic error distribution.

Results

The mean WTP estimate for a program preventing 1 in 10 deaths ranged from $85.16 annually in donations to prevent firearm suicides to $145.63 in additional taxes to prevent deaths from mass shootings. In general, firearm owners were willing to pay less than non-owners; however, Black firearm owners reported the largest WTP, among subgroups and overall. Most respondents were willing to pay the sum-total of their bids to prevent all 3 types of firearm injury; of those, maximum WTP, on average, was $508.08 annually in donations or $534.82 in additional taxes. This implies a statewide total of up to $6.9 billion in perceived benefit.

Conclusion

As resources for prevention, intervention, and supportive services are threatened or terminated, these findings underscore the substantial public demand for investments in firearm injury reduction efforts.

There is substantial public demand for investing in firearm injury prevention. Findings from this general population contingent valuation study in California indicate that most respondents are willing to pay more than $500 annually, on average, for a 10% reduction in firearm-related deaths, implying a statewide estimate of $6.9 billion in perceived annual benefit.

Introduction

Firearm violence and suicide are preventable public health and safety concerns. Following a surge in firearm-related deaths during the early years of the COVID-19 pandemic—when the mortality rate increased by more than 20% from 2019-2021, particularly in Black and Brown communities that have long been disproportionately impacted,¹ and firearms surpassed motor vehicle crashes as the leading cause of death for children and teens—this upward trend has since plateaued and, in many localities across the United States (US), declined to near or below pre-pandemic levels.^2,3 This reversal was not inevitable but resulted instead from deliberate actions, including public and private investments in coordinated, community-partnered firearm injury prevention and intervention strategies grounded in shared values of compassion, care, and systems of support.^4,5

The benefits of these strategies are far-reaching because so too are the harms, which reverberate through networks, neighborhoods, and the nation as a whole.⁶ For every firearm death or injury, the lives and livelihoods of countless more people are indirectly and vicariously impacted, as witnesses or loved ones and through experiences of collective trauma and anticipatory fear stemming from the threat of harm and by actions to avoid and mitigate those risks.^7-22 Enumerating this broader burden of firearm-related harm and, by extension, the value people assign to preventive interventions, necessitates methods for assessing the direct material costs of injury, illness, and crime—eg, medical care, law enforcement, institutionalization, and lost workforce productivity—as well as the less tangible assets that are accrued, at least in principle, from gaining, preserving, or avoiding the loss of sense of safety and other quality of life outcomes.

Contingent valuation is one such method in which survey respondents are asked to quantify their willingness to pay (WTP) for non-market public goods or services. While past research has used contingent valuation to calculate the value assigned to preventing violent crime and suicide,^23-25 less is known about firearm-specific outcomes. One recent exception, an update to a 1998 study, estimated that households are willing to pay an average of $744 annually for a 20% reduction in firearm-involved crime, implying a national total of $97.6 billion in perceived benefit.^26,27 At a time when federal grants that fund the study and practice of firearm violence and suicide prevention, intervention, and supportive services have been threatened or terminated,²⁸ this study provides a monetary estimate of the public's support for, and social return on, investments in firearm injury prevention. Given that the prevalence, perceptions, and experiences of firearm ownership and firearm-related harms are systemically embedded and racially unequal, we assess WTP overall, by firearm ownership status, and among racial and ethnic subgroups of gun owners.

Data and methods

Data

Data come from the 2020 California Safety and Wellbeing Survey, a repeated cross-sectional survey administered online in English and Spanish in July 2020 by the survey research firm Ipsos. Adult (≥18 years) residents of California households in the Ipsos KnowledgePanel were eligible to participate. Of 5018 invitations, 2870 (58%) respondents completed the survey. The survey was approved by the UC Davis Institutional Review Board. Additional information about processes and procedures for the Ipsos KnowledgePanel is detailed elsewhere.²⁹

Measures

We assessed WTP for preventing firearm homicide, firearm suicide, and deaths from mass shootings using the following dichotomous choice (yes/no) question:

“Imagine there is a program that prevents 1 in every 10 [“gun homicides” OR “gun suicides” OR “deaths from mass shootings”]. If this program were available in California, would you be willing to [“pay in extra taxes” OR “donate”] [“$50” OR “$100” OR “$200”] per year to support this program given your other expenses?”

Employing a double-bound approach consistent with prior research (Figure S1),^26,27 respondents who answered “yes” were asked whether they would be willing to pay double the initial dollar amount per year. Those who answered “no” were asked whether they would be willing to pay half. Respondents were assigned the highest value to which they agreed to pay; those who answered “no” repeatedly were assigned a value of $0.

Respondents were asked about the 3 firearm injury types separately to reflect differences in their social construction and public meaning; though, we acknowledge that the different injury types may not be mutually exclusive. The order of injury-specific question presentation, initial dollar value ($50, $100, or $200), and payment condition (TAX or DONATE) were randomized. The same initial dollar value and payment condition were applied across firearm injury types. No information was provided on the epidemiology of firearm violence or suicide. Respondents were therefore assumed to answer based on their pre-existing understanding of firearm-related deaths rather than a standardized accounting of costs and consequences.

To obtain an overall estimate of the value respondents assigned to preventing all 3 types of firearm-related harms and to assess “income effects” associated with adding separate WTP amounts together (eg, a respondent might be willing to pay $50 each to prevent firearm homicide and firearm suicide, but not $100 to prevent both), respondents who said they were willing to pay any amount for at least one type of injury were asked whether they would be willing to pay the sum-total of their valuations “to prevent all of this gun violence” (yes/no); regardless of their answer, respondents were also asked to write in the total amount they would be willing to pay.

Respondents were grouped by firearm ownership status (firearm owners, non-owners living with owners, and non-owners in households without firearms) and racial and ethnic identity. Additional sociodemographic information included age, sex, marital status, household income, household size, and children in the household (Table S1).

Analysis

For each payment condition and subgroup, we fit a double-bounded dichotomous choice contingent valuation model with a log-logistic error distribution, adjusting for sociodemographic variables using a cubic spline to accommodate non-linear effects. For consistency with past research, we present mean WTP after removing (truncating) the highest bid values to address “fat tails” in the dataset. Bootstrap standard errors (10 000 replications) were used to calculate 95% confidence intervals (CIs). The models are fit by minimizing a weighted likelihood function using a survey weight variable provided by Ipsos that adjusts for the initial probability of selection into KnowledgePanel and for survey-specific nonresponse and over/under-coverage. With weighting, the sample is designed to be statistically representative of the non-institutionalized adult population of California. Analyses were conducted in Stata, version 15.1 (StataCorp) and with the DCchoice package (version 0.1.0) in R, version 4.0.0 (R Foundation for Statistical Computing).

Results

The percentage of respondents answering affirmatively to the first WTP question generally declined as the initial dollar value increased, regardless of firearm injury type and payment condition (Figure S2). Even so, a majority of respondents in the TAX condition were willing to pay to prevent all 3 types of firearm-related death, including 57.4% (95% CI: 53.4-61.2) who said “yes” to the first and/or follow-up WTP questions about deaths from mass shootings, 53.4% (95% CI: 49.3-57.3) for homicides, and 49.5% (95% CI: 45.5-53.5) for suicides (Table S2). In the DONATE condition, 46.9% (95% CI: 43.0-50.8) to 50.7% (95% CI: 46.8-54.6) of respondents, depending on the firearm injury type, answered “yes” to the first and/or follow-up WTP questions.

Among respondents in the TAX condition, mean WTP for a program that prevents 1 in 10 firearm-related deaths was $116.14 (95% CI: 106.17-127.03) annually for suicides, $136.37 (95% CI: 125.68-147.67) for homicides, and $145.63 (95% CI: 134.79-156.89) for deaths from mass shootings (Figure 1, Table S2). In the DONATE condition, mean WTP was slightly less: $85.16 (95% CI: 77.85-93.33) for suicides; $93.70 (95% CI: 85.72-102.25) for homicides; and $103.05 (95% CI: 94.74-111.94) for deaths from mass shootings.

Figure 1.

Truncated mean willingness to pay (WTP) to prevent 1 in 10 firearm-related deaths, by firearm injury type and payment condition. The error bars show 95% confidence intervals.

Of those who were willing to pay more than $0 to prevent at least one type of firearm-related death, 69.5% (95% CI: 64.6-74.0) of respondents in the TAX condition and 63.8% (95% CI: 58.5-68.8) in the DONATE condition said they were willing to pay the sum-total of their previously endorsed bids to “prevent all of this gun violence”; of those, maximum WTP per year, on average, was $534.82 (95% CI: 408.85-660.79) in the TAX condition and $508.08 (95% CI: 405.97-610.20) in the DONATE condition (Table S2). Among those who were not willing to pay the sum, maximum WTP, on average, was $156.42 (95% CI: 117.87-194.96) annually in the TAX condition and $133.62 (95% CI: 110.52-156.72) in the DONATE condition.

Firearm owners and non-owners living with owners endorsed a smaller mean WTP than non-owners in homes without firearms, regardless of payment condition; these differences were more pronounced in the TAX condition (Table S3). When firearm ownership was disaggregated by race and ethnicity, different patterns emerged (Figure 2, Table S4). Of particular note, Black firearm owners endorsed larger WTP than Black non-owners in homes without firearms, across all firearm injury types and payment conditions (depending on firearm injury type and payment condition, $168.04 [95% CI: 93.83-256.65] to $305.39 [95% CI: 140.36-384.42] among Black firearm owners vs $59.68 [95% CI: 36.47-92.64] to $140.80 [95% CI: 102.59-185.80] among Black non-owners in homes without firearms); WTP among Black firearm owners was the largest of any ownership subgroup and overall.

Figure 2.

Truncated mean willingness to pay (WTP) to prevent 1 in 10 firearm-related deaths among firearm owners, by race and ethnicity, firearm injury type, and payment condition (tax [A] or donate [B]). The error bars show 95% confidence intervals.

Discussion

Firearm violence and suicide have far-reaching consequences that include devastating losses of life and life-altering injuries, but also myriad societal spillover effects. A conventional “cost of injury” (COI) approach to calculating this burden often involves totaling the dollar amount of medical costs, lost earnings, and in some versions, the “subjective value” of the lives of those killed or injured; essentially, the monetary equivalent of counting only the costs accrued to direct victims. The contingent valuation approach attempts to more broadly encompass the indelible marks of firearm-related harms on our shared consciousness and daily lives, thereby facilitating an expanded conceptualization of the nature of this burden and, by extension, the public good of preventive interventions in service of improved population health and safety.³⁰

WTP estimates from this study are similar but not directly comparable to those from past research on firearm-related harms,^26,27 and those from past research on other endemic public health topics, including a WTP estimate from November 2020 of $318.76 for a hypothetical COVID-19 vaccine with high efficacy.³¹ In addition to differences in question wording, dollar amounts, and proposed percentage reduction, we asked separately about types of firearm injury, including firearm suicide, and varied the payment mechanism. Depending on the firearm injury type and payment condition, we estimated that the mean WTP to prevent 1 in 10 deaths ranges from $85.16 annually in donations to prevent firearm suicides to $145.63 annually in additional taxes to prevent deaths from mass shootings.

Notably, Black firearm owners endorsed the largest WTP. Alongside prior evidence indicating greater support for gun safety policies and community-based gun violence prevention efforts,³² and also increased awareness of inequities in gun injury,³³ among Black firearm owners nationally, our findings indicating majority support for investment in prevention programming, particularly among this subgroup, may be further evidence of racially divergent feelings of safety resulting from firearm acquisition,³³ despite being broadly motivated by self-protection,³⁴ and the lived experiences of communities of color, particularly Black communities, of being overpoliced but underserved by conventional systems of public safety.³⁵

Across subgroups, most respondents were willing to pay the sum-total of their bids for all 3 firearm injury types; of those responding affirmatively, maximum WTP, on average, was $534.82 annually in additional taxes or $508.08 in donations. Applying a simple extrapolation that assumes 12 965 435 households in California and a conservative assumption of only one contributor per household, our estimates therefore imply that the statewide benefit of a 10% reduction in firearm-related deaths is $6.6-$6.9 billion per year, more than double the estimated $3.0 billion in savings for a 10% reduction based on a COI approach that determined firearm-related deaths cost Californians $30.2 billion per year.³⁶ As just one point of reference, based on a recent calculation that jurisdictions should anticipate spending between $750,000 and $4 million annually on community violence prevention programming,³⁷ this $6.6-6.9 billion estimated valuation would support at least 1650 such programs across California's 482 municipalities. That WTP in additional taxes was nearly always higher than WTP in donations additionally suggests that respondents perceived the prevention of firearm injury not only as an individual task but as a public responsibility.

This study has limitations. First, we use self-reported survey data, which is subject to social desirability bias and may lead to overestimates of WTP. Second, respondents were not asked about all types of firearm injury (eg, unintentional) nor provided with information about the epidemiology of firearm deaths. Details about the proposed preventive measures were also omitted, which may constrain confidence in the stated reduction, or respondents may object to firearm-related programs for other reasons. Together, these factors may contribute to underestimates. Third, although we think these findings can inform understandings of the demand for preventive interventions in other states and nationally, differences in the cultural acceptability of firearms, more restrictive firearm regulations, and lower rates of firearm ownership and firearm-related deaths in California should be considered when attempting to generalize these results. Finally, data collection for this study occurred during the COVID-19 pandemic when many people experienced job losses and other financial strains, which may have led to more conservative WTP estimates.

Conclusion

Preventing firearm-related harm is a common good shared by all those whose lives and livelihoods benefit from healthier and safer communities. At a time when recent reductions in rates of gun-related morbidity and mortality are threatened by rollbacks in funding for policy, research, and community-led strategies that address the causes and consequences of firearm violence and suicide, this analysis underscores the substantial public demand for and valuation of investing in comprehensive firearm injury prevention strategies.

Supplementary material

Supplementary material is available at Health Affairs Scholar online.

Conflicts of interest

The authors have declared that no competing interests exist.

Please see ICMJE form(s) for author conflicts of interest. These have been provided as supplementary materials.