Some Thoughts on Quantifying Death

I’ve been thinking a lot about death recently. Emotional responses to a couple recent events that directly affected me triggered these thoughts. The first was a recent wave of urban cyclist deaths in US cities, most notably two hit-and-run killings on June 22nd within hours of each other on the bike paths I ride daily in San Francisco. The second was the Alton Sterling and Philando Castille killings on July 5th and 6th respectively and the subsequent social unrest.

Death is natural, every one of us will experience it, and roughly 7,100 people die in America every day but these deaths mentioned above struck a chord with me. These people didn’t live long, productive lives before dying of natural causes. They did not know their time was coming and could not have anticipated that their lives would come to an abrupt end on those days they deceased. They were going about their business and another member of our society, whether through conscious action or careless inaction, quickly and abruptly ended their time with us. Roughly 7,100 people die every day in America but deaths are not equal and the nature of these deaths stood out quite alarming to me.

If not all deaths are equal, what differentiates one type of death from another? Personally, I tend to evaluate a death on two measures:

1. Type of Death

While these are broad, generalized, subjective categories that can have many overlaps, I tend to classify “Type of Death” in three ways:

  1. Death from natural causes — This includes death due to old age (when we should die) as well as other diseases that can unexpectedly strike at younger ages like cancer.
  2. Death stemming from our own behavior — This includes self-inflicted death like suicide as well as deaths resulting from the consequences of engaging in risky behavior (Bungee jumping off cliffs, having frequent unprotected sex with random strangers, eating donuts every day, etc)
  3. Deaths caused by other members of society — This could be through an individual’s active decision that someone else should die (murders) or their careless and/or negligent behavior (texting while driving).

While the lines for these three groups are extremely vague (is death stemming from heart disease at age 80 a natural cause or does it stem from my own behavior if I ate donuts every day? Is death caused by someone else texting a death stemming from my own decision to engage in the risky behavior of driving on our nation’s roads with other careless individuals?), categorizing these helps me think more clearly about the different ways people die and the decisions we can make to work towards reducing deaths.

2. Years Lost

Everyone will die at some point. Many people will die of natural causes at an old age. My personal believe is that a person who lives a long life and dies of natural causes at an old age represents a healthy, functioning evolutionary progression. They have lived how their bodies and minds were developed to live. Pushing it any longer would be a waste, cutting it any shorter would be a loss. Therefore, anyone who doesn’t make it to this hypothetical “working as intended death from a natural cause in old age” has lost years and all lost years are created equal.

Of course, this is an extreme simplification and a seemingly unlimited number of factors can go into weighing the value of years lost (are years lost from a nobel laureate in their academic prime who dies at 43 more valuable than years lost from a convicted killer serving a life sentence in prison for heinous crimes who dies at 42?) but my goal is to keep things simple and think about the inequality of death on very basic, simple, and easily measurable terms.

Thinking about “Type of Death” and “Years Lost” as classifications for death, I realized that the recent deaths that bothered me emotionally tended to fall into the categories of “death caused by other members of society” that resulted in a high number of years lost. To me, these represent both a great loss given the number of years of life our society loses as well as a strong individual questioning of whether I, as a member of my society, took enough measures to protect someone else in my society from a death that another society member inflicted upon them. In the recent cases above, I feel that I have not and I feel a great sense of guilt and remorse that many years were lost, and continue to be lost, as a result of actions that myself and my society are not working hard enough to prevent.


Curious about the frequency of such deaths, I poked around at the NVSS Mortality Data published by the CDC to see if I could find any interesting signals that highlighted the severity of these death classifications I had been bouncing around in my head. (Disclaimer: this is not an academic paper with thoroughly tested and peer reviewed analytical rigor but rather a casual afternoon investigation looking into some basic trends that intrigued me that I finally got around to posting about. If you are curious about my analysis and methodology, let me know and I’d be happy to share).

The CDC uses standard classifications for mortality with tens of thousands of unique cases but in order to simplify the categories, I took the CDC data (Compressed Mortality File 1999–2014 Series 20
No. 2T, 2015, as compiled from data provided by the 57 vital statistics jurisdictions through the Vital Statistics Cooperative
Program) and grouped all deaths into the following categories that made sense to me:

  • Cancer
  • Heart Disease
  • Other Diseases
  • Auto Deaths
  • Gun Homicides
  • Other Homicides
  • Accidents
  • Suicides
  • Other Deaths

They also put deaths into age buckets (e.g. ages 1–4, 5–9, etc) which I used to calculate a rough estimate of “years lost” buckets for each death by subtracting the median age for each age bucket from the US life expectancy of 79 years (Note: I excluded all deaths that were above the median age for these calculations to avoid zero or negative years lost and excluded deaths between ages 0–1 because many involved accidents or disease at birth which created lots of noise in the data).

Once I calculated a “Years Lost” bucket for each death, I calculated a weighted average of years lost for each category listed above. Below are a few charts (click on them to magnify).

Chart 1 — % of Deaths per Age Group and Category (excludes Other Homicides, Accidents, Suicides, Other Deaths)

Chart 2 — % of Deaths per Age Group and Category (Includes all categories)

Chart 3 — Years Lost per Category (weighted average)

Points that I found interesting

  1. If you survive until age 45, chances are you will die from Heart Disease, Cancer, or Other Diseases (chart 2 shows this well)
  2. Auto Deaths and Gun Homicides make a up a noticeably high percentage of deaths among 15–24 year olds (chart’s 1 and 2)
  3. The average years lost from a Gun Homicide is extremely high (chart 3)

Addressing point 1, this may seem quite obvious but we are all going to die at some point and if we live long enough, we will probably die of natural causes (e.g. cancer, heart disease, other diseases). This is totally normal and I don’t think we as a society should place a great deal of emphasis on old people dying either from a financial standpoint nor from a data and publicity standpoint. I’m sick of hearing people and the media state that diseases old people die from are the “top killers in America” because everyone is going to die at some point and old people dying from diseases is for the most part completely natural and, in my opinion, should not be a primary focus of our society. In fact, I think we should embrace the fact that we will reach death in old age and make some sacrifices for society by pulling our plugs a few days sooner. A shocking amount of our nation’s health care costs (and tax dollars) pay for a patient’s healthcare in their last several months of life and could probably be spent in better ways (feel free to disagree!).

Regarding points 2 and 3, I believe that our country and society should expend a great deal more effort and resources on preventing deaths that both result in many years lost and can be quickly and easily caused by other members of society. Auto and gun homicide deaths stand out among the rest of these categories in this regard. Unlike elderly people dying because they have reached their time, auto and gun deaths shock (or should shock) our society because these often involve deaths caused by other members of society and we can attribute a high percentage of young people’s deaths to these (many auto deaths are probably self-inflicted but not by means of a risk-calculated choice like eating donuts every day but rather by momentary carelessness). When these types of deaths occur and we become exposed to them, we feel guilt and pain because we, as a society, have not taken the appropriate measures to prevent them, we know that these should not occur, and we know that we have let our people down.

If we see high rates of young people dying in these ways, we should attack these categories like a virus; like we would work to contain a smallpox epidemic, an ebola outbreak, or a case of Zika virus. Today’s New York Times headline read “New York City Wages War on the Zika Virus” with a bold picture of emergency vehicles patrolling the streets but why is New York City not publishing stories about waging a war on the systemic social issues that lead to the steady stream of its auto and gun-related deaths like the drive-by double shooting that killed 17 year old Dontay Gordon in the Bronx last Wednesday (62 years lost). By the way, I randomly googled “shooting nyc” while writing this and found that article about Dontay Gordon and, to be completely honest, I wasn’t surprised that the first article I found was about a young, black male shot in a drive-by shooting in the Bronx, which means there is something horribly broken in our society at the moment and we should be doing much much more to fix this. What’s more upsetting is that I googled his name and he’s the 2nd son the family has lost to gun violence. How do we let this happen and why are we not waging a war against this?

And why do we let intoxicated drivers hop into vehicles and carelessly mow down cyclists like this Google software engineer and father of three who was run over and killed by a drunk 24 year old girl in June? I also found this by googling “drunk driver cycling death” while writing this. All deaths are tragic but when a 35 year old dad riding his bicycle is accidentally run over and killed by a drunk driver, the incident devastates the life of the driver as well as the victim’s three young sons, his wife, and his countless friends and colleagues.

It should not be so easy for one member of society to end and ruin so many lives, whether careless or planned, in one brief moment. We should do everything in our power to make it very difficult for any random member of society to control a machine that can quickly and easily end other people’s lives. Nevertheless, our nation’s fatalities clearly indicate that the importance we place on our freedom to own and operate automobiles and guns allows us to end other people’s lives with ease and we end up destroying many young people’s lives as a result.

Like we fight diseases and wars, we should wage an all-out war on the current systems in place that allow for so many auto and gun-related deaths. We should home in on each and every death that occurs from vehicles and guns and devote every resource we can to stamping out the culprits at the source. We should continually track and analyze and scrutinize each mortality statistic and invest the resources necessary to eradicating gun and auto fatalities. I am certainly not an expert on gun control nor automobile-related fatalities but the alarming rate at which these occur leads me to believe that we can do much better than we are currently doing. Maybe this involves safer street plans for cyclists and pedestrians, breathalyzers for car ignitions, free transportation from bars, greater investments in autonomous vehicle technology and legislation, improved firearm background checks and regulation, or improved ammunition regulation like California is doing.

I do not know the answers but I do continue to see these deaths occurring all around me and I feel responsible and guilty that I am letting my society down as a result. I hope that many political candidates identify and focus on these issues during the upcoming election season. In the meantime will try to make a concerted effort to do what I can to make the streets of San Francisco safe for pedestrians and free from gun violence.

I hope that, as our technological capabilities and data collection practices improve, we can use these to more effectively analyze each and every death that occurs and actively invest in efforts to reduce future deaths so that everyone can live a long, productive, and happy life free from the surprises of stray bullets and swerving cars. When we eventually eradicate or greatly contain gun and auto deaths (which I am confident we will do during my lifetime), I will glance back at these numbers again some afternoon and write a blog post on the social ills of daily donut consumption :)