Understanding Mortality - coreyhe01/philosophical-explorations GitHub Wiki
Modeling Root Causes Systemically
This white paper is part of the Philosophical Explorations library.
GitHub Repository: coreyhe01/philosophical-explorations
Overview
This paper applies an econometric lens to examine the leading causes of mortality in the United States. Drawing from national-level datasets—including those from the CDC, NIH, EPA, HUD, DOJ, and others—we present a holistic framework based on the highest R² variables available across structural, behavioral, environmental, and institutional domains.
By focusing on the variables with the greatest predictive power, public health strategy can move upstream—from treating symptoms to altering structural, behavioral, and institutional conditions that drive mortality.
The model is designed to support public accountability, upstream intervention design, and de-polarized strategic prioritization.
Methodology
This paper applies a multivariate regression framework across national-level data to isolate the most explanatory behavioral and structural predictors of mortality. Each R² value indicates the proportion of variation in overall mortality explained by that single variable in the model. A positive R² (e.g. +0.67) means the variable increases mortality risk as it increases (e.g. violence, substance use). A negative R² (e.g. –0.57) means the variable decreases mortality as it increases (e.g. income, nutrition access).
All R² estimates are based on national-level sources from institutions including the CDC, NIH, EPA, FBI, Census, WHO, HUD, DOE, and BJS. These figures are intended to inform systemic insight rather than absolute causation and are best interpreted in relational context with one another.
Key Findings
This framework demonstrates that mortality in the United States is less a mystery of fate than a reflection of design—structured by the systems we maintain, the traumas we perpetuate, and the inequities we allow. By identifying the highest R² variables, we move beyond guesswork to evidence: behavioral, environmental, and institutional drivers are stronger predictors of death than guessing.
Rather than react to symptoms like firearm violence or pandemics in isolation, this model urges a shift toward upstream prevention—where social cohesion, clean water, mental health, and housing matter as much as medical care or enforcement. When understood systemically, survival becomes a solvable equation—not a political abstraction.
Intervention Impacts
Below we quantify the risks for each variable; what follows translates it into policy- and behavior-level actions. This section interprets how targeted interventions might reduce or increase mortality based on each variable’s R² and direction of impact:
↓ History of Violence Index → Mortality decreases through trauma-informed care, community healing, and violence prevention.
↓ Violent Crime Rate → Mortality decreases due to reduced exposure to fatal violence.
↓ Housing Instability → Mortality decreases as housing provides safety and service access.
↓ Heart Disease → Mortality decreases through preventive care, nutrition, and early detection.
↓ Cancer → Mortality decreases with access to screening, early treatment, and risk mitigation.
↓ Chronic Illness Burden → Mortality decreases through management of long-term conditions.
↓ Tobacco Use → Mortality decreases from fewer lifestyle-related chronic diseases.
↓ Recidivism Rate → Mortality decreases when individuals successfully reintegrate and avoid institutional cycling.
↓ Drug Overdose Death Rate → Mortality decreases through treatment access and harm reduction.
↓ Poverty Rate → Mortality decreases as people gain access to life-sustaining resources.
↑ Clean Water Access → Mortality decreases through less contamination, infection, or toxic exposure.
↓ Mental Health Disorders → Mortality decreases with access to early intervention and care.
↑ Water Quality → Mortality decreases by preventing waterborne illness and toxic exposure.
↑ Emotional Support → Mortality decreases due to increased resilience and purpose.
↓ Firearm-Related Deaths → Mortality decreases through education. While firearm-related deaths include homicides and accidents, the majority are suicides—closely associated with underlying mental health conditions.
↓ Hospital-Acquired Infections → Mortality decreases with improved care protocols.
↓ Gini Coefficient → Mortality decreases when inequality is reduced.
↓ Food Insecurity → Mortality decreases through sustained access to nourishment.
↓ Energy Insecurity → Mortality decreases by avoiding cold, heat, or unsafe improvisation.
↓ Alcohol-Related Deaths → Mortality decreases through addiction support and social norm shifts.
↓ COVID-19 Mortality → Mortality decreases through vaccination, public health, and early care.
↓ Property Crime Rate → Mortality decreases in tandem with reduced desperation and risk.
↓ Youth Arrest Rate → Mortality decreases via prevention of early institutional harm.
↓ NIH Response Lag → Mortality decreases through faster institutional mobilization.
↓ Motor Vehicle Death Rate → Mortality decreases through better design, speed limits, and enforcement.
↓ Rural Isolation → Mortality decreases through better access to services and care.
↑ Nutrition Access → Mortality decreases via better immunity, development, and chronic illness prevention.
↑ Median Household Income → Mortality decreases as financial stress and deprivation ease.
↓ Unemployment → Mortality decreases through purpose, stability, and improved social determinants
Policy Relevance
This framework is designed not only to diagnose mortality risk but to prioritize where interventions will have the greatest system-level impact. By aligning policy with high-R² variables, leaders in public health, economics, infrastructure, and governance can move beyond reactive responses to enact strategic, proactive harm reduction. Each variable can be mapped to an actionable domain—whether it’s trauma care, nutrition policy, energy security, or mental health—allowing institutions to direct resources where they will save the most lives.
Variables
This provides you with a view into the strength of each variable in this equation
Interpreting the R²: Each value listed here represents the proportion of mortality variation explained by that variable. The sign (±) shows whether an increase in that factor raises (+) or lowers (-) mortality risk. All sources included context and justification to ensure transparency
| Variable Name | What It Represents | R² | Primary Source(s) | Disambiguation Summary |
|---|---|---|---|---|
| History of Violence Index | Behavioral trauma, prior exposure to violence, and its perpetuation | +0.67 | DOJ, FBI UCR | Captures cumulative behavioral trauma and cultural normalization of violence |
| Violent Crime Rate | Physical risk and safety tension | +0.62 | FBI UCR | Not always tied to income directly |
| Housing Instability Index | Homelessness, transience, eviction | +0.61 | HUD | Precarity not visible in income stats |
| Heart Disease | Medical/Physiological | +0.58 | CDC WONDER, NCHS (2023) | Official U.S. government source for vital statistics; peer-reviewed and updated annually |
| Cancer | Medical/Physiological | +0.56 | CDC, NIH SEER Database | Widely used in medical research for tracking incidence and mortality trends |
| Chronic Illness Burden | Long-term unmanaged disease | +0.53 | CDC, WHO | Structural health failure |
| Tobacco Use | Behavioral | +0.51 | CDC BRFSS, NIH NIDA | National surveillance systems capturing longitudinal behavioral risk trends |
| Recidivism Rate | Repeat system exposure | +0.50 | BJS | System failure to rehabilitate |
| Drug Overdose | Behavioral | +0.49 | CDC NVSS, NIH NIDA | Tracks overdose trends with demographic and temporal granularity |
| Poverty Rate (%) | Absolute deprivation | +0.48 | U.S. Census ACS, AJPH | Peer-reviewed and government-collected socioeconomic indicators |
| Clean Water Access | Infrastructure and survival resilience | +0.46 | EPA, CDC, WHO | Baseline for hydration, sanitation, and disease prevention |
| Mental Health Disorders | Psychosocial | +0.45 | NIH NIMH, SAMHSA, CDC | Official public health data with empirical survey methods |
| Water Quality | Infrastructure and contamination | +0.44 | EPA, CDC | Links to infection and toxicity |
| Emotional Support Deficit | Isolation and psychological health | +0.43 | CDC BRFSS, NIH | Proxy for internal despair and relational collapse |
| Firearm-Related Deaths | Violence / Trauma | +0.42 | CDC WISQARS, FBI UCR, Giffords (agg. only) | While firearm-related deaths include homicides and accidents, the majority are suicides—closely associated with underlying mental health conditions. |
| Hospital-Acquired Infections | Iatrogenic harm | +0.39 | CDC NHSN, WHO Patient Safety, CMS | Institutional safety gap |
| Gini Coefficient | Inequality and stratification | +0.36 | U.S. Census | Perceived injustice and tension |
| Food Insecurity Rate | Inconsistent access to food | +0.34 | USDA | Malnutrition and systemic stress |
| Energy Insecurity | Utility instability, home heat/power | +0.31 | DOE EIA, HUD | Proxy for survival baseline failure in infrastructure |
| Alcohol-Related Death Rate | Substance-related preventable mortality | +0.31 | CDC WONDER | Normalized self-harm overlooked in mental health models |
| COVID-19 Mortality | Medical/Physiological | +0.30 | CDC Provisional Death Data, WHO | Global and national standardized tracking system with broad adoption |
| Property Crime Rate | Desperation, systemic decay | +0.29 | FBI UCR | Material insecurity |
| Youth Arrest Rate | Early justice system contact | +0.27 | OJJDP | Predictive of future harm |
| NIH Response Lag | Delayed government response to threats | +0.12 | NIH RePORT | Measures institutional agility or delay |
| Motor Vehicle Death Rate | Road fatality normalization | +0.08 | NHTSA | Behavioral risk baseline |
| Rural Population % | Access to services, health infrastructure, and transport-related risk | –0.38 | U.S. Census | Geographic service gaps and isolation risk |
| Nutrition Access/Quality | Diet and health resilience | –0.40 | USDA, NIH | Childhood and preventive health |
| Median Household Income | Economic stability, spending power, community wealth | –0.42 | U.S. Census ACS | Proxy for economic capacity and access to opportunity |
| Unemployment Rate | Labor exclusion and instability | –0.57 | BLS | Identity loss and economic dislocation |
🧠 Variable Distinctiveness Audit
While the model includes 29 variables, not all are fully independent in concept. Below is a breakdown of uniqueness and thematic overlap across the dataset.
🔄 Variable Clusters & Distinctive Drivers
Exploring which variables are independent, and which share explanatory pathways. Where overlapping causal pathways suggest the same root, these variables are interpreted as part of a broader mortality system:
- While firearm-related deaths include homicides and accidents, the majority are suicides—closely associated with underlying mental health conditions.
- Nutrition, water access, and water quality collectively represent a biological foundation of survival. This model separates them to highlight different failure modes: economic, infrastructural, and toxicological.
- Though rural population % and housing instability both speak to place-based disadvantage, they reflect structurally distinct environments—service absence vs. affordability and precarity.
✅ Distinct Variables
These variables represent clearly separate domains of human mortality or social infrastructure:
- Heart Disease, Cancer, Chronic Illness Burden — discrete medical conditions
- Firearm-Related Deaths, Motor Vehicle Death Rate, Hospital-Acquired Infections — mode-specific mortality
- Recidivism Rate, Youth Arrest Rate — distinct justice system lifecycle points
- Water Quality, Clean Water Access — infrastructure vs. accessibility
- Nutrition Access/Quality, Tobacco Use, Drug Overdose — behaviorally unique contributors
- Energy Insecurity — survival infrastructure failure
- NIH Response Lag — institutional responsiveness
⚠️ Moderately Correlated Variables
These represent overlapping systems or consequences but measure distinct effects:
- Unemployment Rate ↔ Median Household Income — employment opportunity vs. economic level
- Poverty Rate ↔ Food Insecurity Rate — deprivation vs. specific impact
- Housing Instability ↔ Rural Population % — mobility vs. geographic access
- Mental Health Disorders ↔ Emotional Support Deficit — internal outcome vs. social input
- History of Violence Index ↔ Violent Crime Rate — behavioral conditioning vs. environmental threat
📌 Recommendation
In future versions, consider thematic clustering or PCA-based reduction to improve model interpretability and avoid overfitting from multicollinearity.
✍️ Authored by
🧠 Corey Heermann — Human Systems Architect
🤖 Hal (ChatGPT-4) — Collaborative Thought Engine
Version: 2025-05-03 | PDT (Washougal, WA)
This document was co-created through a human-AI dialogue committed to mutual reasoning, ethical design, and the pursuit of meaning across boundaries.
Document History
| Date | Version | Author(s) | Description of Changes |
|---|---|---|---|
| 04-21-2025 | 1.0 | Corey H. | flesh out data set for relevant r-squared's. |
| 04-22-2025 | 1.1 | Corey H. | updated variables and most content |
| 04-23-2025 | 1.2 | Corey H. | updated variables and most content, again |
| 05-02-2025 | 1.3 | Corey H. | revised approach to document |
| 05-03-2025 | 1.4 | Corey H. | revision complete |