Internet search patterns reveal firearm sales, policies, and deaths

Firearm-related violence is a major source of morbidity and mortality in the United States, with an average of 36,538 deaths and 100,102 injuries per year between 2013 and 2017. High-profile mass shootings further fuel the ongoing debate on firearm policy. Central to this debate is the challenge of understanding the impact of policy on firearm ownership and firearm-related morbidity and mortality. Despite the critical role of understanding these policy outcomes, accurate and comprehensive data on gun sales and ownership are not available. Current research often forms policy insights on data derived from surveys, proxy variables, or production and import data. The most common data source used in firearm policy research is the Federal Bureau of Investigation’s National Instant Criminal Background Check System (NICS). While this information is a useful surrogate measure, it does not represent actual firearm sales and is greatly impacted by regulations that vary from state to state. In some cases, NICS data may represent an overestimation due to permit denials, multiple background checks conducted for single firearm purchases, or waiting periods that deter eventual firearm sales. In other cases, NICS data may underestimate firearm sales due to multiple firearm purchases for a single background check, exemption from background checks based on concealed handgun permits, and lack of information on sales by private sellers, including those conducted at gun shows. Federal provisions that limit certain agencies from engaging in gun control research and tracking have further hindered accurate firearm surveillance. The central importance of the gun control debate in the public sphere, together with the current restrictions on data-gathering, drive the need for alternative, low-cost and timely sources of firearm-related data. In recent years, a new generation of public health surveillance efforts has relied on patterns of internet searches. Search data can be used to synthesize population health and evaluate the impact of health policies. The present study extends previous work by conducting an in-depth analysis of gun-related search volumes and their relationships to different gun-related phenomena. We compare search query volume for a range of general and gun-type-specific search terms against background-check data over both time and space, stratify analyses by gun type, and examine relationships to state-level firearm-related mortality and state-level firearm policy. Finally, we propose a framework for prospective surveillance that incorporates gun-related internet search volume as a freely available, real-time complementary data source that enables analysis at finer geographic and temporal scales, with fewer delays in data collection, to inform public policy.

Prior research has demonstrated the utility of internet search data for public health surveillance, including applications in infectious diseases, abortion, smoking, and mental health. Gun-related internet search studies have shown that searches for terms like “bury gun” correlate with firearm background checks (e.g., 2008–2013, Pearson’s r = 0.84) and that firearm-related search volumes spike following mass shootings, with increases of tens to hundreds of percent. Recent work also reported increased gun-related searches during the COVID-19 pandemic. These studies primarily focused on temporal spikes and single-term correlations. The current study builds on this by systematically assessing correlations across time and space, stratifying by gun type, and linking search data to firearm mortality and policy restrictiveness.

Data retrieval: The study compiled state-level data from multiple sources. Background-check totals (overall, long-gun, handgun) for 2019 were retrieved from the FBI’s National Instant Criminal Background Check System (NICS). State population estimates came from the U.S. Census to compute per-100,000 rates. Firearm-related deaths for 2017 (most recent available) were obtained from CDC WISQARS. State-level firearm policy restrictiveness scores for 2019 were taken from the Giffords Law Center Annual Gun Law State Scorecard (1–50 scale). Internet search volume data were retrieved from Google Trends for a range of firearm-related terms (e.g., gun, shotgun, rifle, pistol, 9 mm). All data were retrieved on February 10, 2020, using the most recent available year for each analysis. Data analysis: The authors conducted correlation analyses (Pearson’s r and Spearman’s ρ) between Google Trends relative search volumes and: (a) background checks over time (monthly U.S. totals in 2019) and across space (state-level per 100,000 in 2019), with stratification by gun type (long-gun vs. handgun); (b) firearm-related deaths across states (2017); and (c) restrictiveness of state-level firearm policies across states (2019). A time-lagged correlation analysis assessed lags from −6 to +6 months between search volumes and background checks to identify lead/lag relationships. Analyses highlighted specific search terms with the strongest associations (e.g., shotgun for long guns, 9 mm for handguns) and summarized results across all terms.

• General term correlation: Search volume for “gun” vs. total background checks in the U.S. showed Pearson’s r = 0.74 and Spearman’s ρ = 0.62 (P = 0.006). - Stratification by gun type markedly improved correlations. Time (U.S. monthly, 2019): long-gun checks vs. “shotgun” r = 0.96, ρ = 0.94 (P < 0.001); handgun checks vs. “9 mm” r = 0.97, ρ = 0.94 (P < 0.001). Space (states, 2019): long-gun checks per 100,000 vs. “shotgun” r = 0.78, ρ = 0.76 (P < 0.001); handgun checks per 100,000 vs. “9 mm” r = 0.63, ρ = 0.59 (P < 0.001). - Time-lag analysis (−6 to +6 months) indicated peak correlations at zero-month lag for both long-gun/shotgun and handgun/9 mm comparisons. - Firearm-related deaths vs. search volumes (states, 2017): “shotgun” r = 0.71, ρ = 0.68 (P < 0.001); “9 mm” r = 0.87, ρ = 0.90 (P < 0.001). - Policy restrictiveness vs. search volumes (states, 2019): strong negative associations observed for “9 mm” (r = −0.82, ρ = −0.83, P < 0.001). The study reports strong associations overall between higher firearm-related search interest and less restrictive firearm policies. - Overall, search query volumes provide a near-real-time signal closely tracking background checks and relating to firearm mortality and policy differences across states.

The study demonstrates that firearm-related internet search query volumes strongly track firearm background checks across both time and space, with the strongest correlations at zero lag, suggesting searches and purchases occur contemporaneously. Stratifying by gun type (e.g., long guns vs. handguns) substantially increases concordance with corresponding background checks, indicating that targeted term selection captures distinct market segments. Extending beyond sales proxies, the analyses show that higher firearm-related search volumes are associated with higher state-level firearm mortality rates and with less restrictive firearm policies, highlighting their relevance for public health and policy surveillance. Given that search data are freely available, near-real-time, and provide fine spatial and temporal resolution, they can complement existing, delayed, or incomplete data sources (e.g., NICS not capturing private sales). Incorporating search-based surveillance could help monitor shifts in firearm interest/use, evaluate impacts of policy changes and public health initiatives, and inform timely interventions.

Internet search patterns offer a valuable, timely, and complementary resource for monitoring firearm-related behaviors and policy environments. By validating search volumes against background checks and demonstrating associations with firearm mortality and policy restrictiveness, the study proposes a framework for prospective surveillance leveraging search data at finer geographic and temporal scales. Future research should: (1) conduct further validation studies linking searches to direct measures of sales and ownership; (2) examine more granular policy components and their relationships to specific search terms; (3) explore sub-state spatial analyses and higher-frequency (weekly/daily) temporal dynamics; and (4) refine term sets to disentangle motivations (e.g., purchasing vs. information-seeking following media events).

• No gold standard exists for firearm sales/ownership; NICS background checks are an imperfect proxy (can over- or under-estimate actual sales; exclude private sales). - Internet access and usage vary by demographic, socioeconomic, and geographic factors, limiting representativeness. - Search intent is unknown; spikes may reflect media coverage or non-purchase interest (e.g., after mass shootings). - Potential sensitivity to outliers affects Pearson’s r; reporting both Pearson’s and Spearman’s mitigates but does not eliminate this concern. - Results depend on term selection; multiple terms and stratification help specificity but residual ambiguity remains. Additional validation is needed to link search behavior to real-world purchasing and ownership patterns.