Counting the Dead - The Killing Age

Counting the Dead

Most of us have all grown up learning about the catastrophes of the twentieth century: WWI and WWII, Stalin, Mao, for example. (On the other hand, surveys have suggested just how little many people know about the past.) This period was, in the words of one famous historian, an “age of catastrophe.” (For more on the twentieth century see my other essays on this website, especially “Numbers and their Discontents,” and “What About Violence in the Twentieth Century?”.)

We have survivors of those catastrophes to remind us of their horrors. They are also part of the public cultures of many areas of the world, especially the West, which has dominated the production of knowledge. And, in some parts of the world, they are part of the historical legitimation of nation-states and nationalist claims, often defined in narrowly ethnic ways.

No one is living who was born before about 1900. And given the emphasis on the West, it is highly likely that a person living in Berlin or Los Angeles—even a highly educated person, would know what unfolded in Bunyoro in the late nineteenth and early twentieth centuries, or even where Bunyoro is (Uganda). Or for that matter Indonesia. Tourists to Mexico’s Yucatan are largely unaware of the bloody Caste War. When you look at the database you are very likely to encounter many places across the world you have never heard of.

An important goal of The Killing Age was to return to this earlier period and to offer a global history of violence the period c. 1750-c.1900. One argument was that this period which saw the making of the modern world, was colossally destructive, for humans and non-humans alike. This means that we have to think about the short twentieth century somewhat differently, as well as the more recent past.

Here is a graph that helps illustrate the point, based on the Brecke databases. There are various challenges/issues in with this database and one really should be most interested in the trends. One challenge–and an important one–is that it struggles to address warlordism, that is non-state violence.

In addition, I have included the general dates for the British and USA industrial revolution, as well as WWI and WW2.

The major point is the rise of conflicts during the industrial revolutions, followed by a general decline with the exceptions of the two world wars. Here is the graph:conflicts_A_1700_1975_annotated

and

conflicts_started_per_decade_1700_1900_line.

I never intended the book to be comprehensive. I was mostly interested in the relationship between violence and capitalist development, or what I called capitalism’ “big bang.” And I offered numbers warning the reader that these were—in some cases could only be—estimates.

Given some of the criticisms the book has received, I decided to go back to the drawing table, to provide some additional statistical details. I did so without going back to the book; as an experiment I thought this would be a good way to see if how far off I might be. Indeed, I haven’t even cross-checked with the book. So this database is a kind of independent exploration.

I have prepared the database for a few reasons. First, as I mentioned some readers seem surprised by the general claim that the period under study could have been so dreadful. Second, the nineteenth century especially seems to be all the current rage. President Trump explicitly invokes, indeed nearly worships, the period. (See the administration’s National Security document, for example.) The nineteenth century has re-entered public culture. Needless to say, as many commentators have noted there are some rather dangerous aspects to the “Donroe Doctrine,” not to mention its historical amnesia.

Interestingly, the totals are remarkably similar to my crude estimates in the book: as of March 23, 2026 writing, the totals stand at 156,669,315 as a reasonable sum, and 287,185,050 for the upper estimate. Here is a graph that helps illustrate the developments: killing_age_deaths_by_decade

One crucial thing to note is the upsurge during the second half of the nineteenth century during a period of rapid European and American imperial expansion.

A second, and broader, important observation is that much of the violence in the twentieth century was the result of the pursuit of political religions (communism, nationalism, and so on), whereas the violence I explore in the book is tied to economic motives. This has begun to change, at least in some areas, most obviously in the DRC where violence is often tied to securing access to scarce minerals.

Obviously, far more work must be done. There are both empirical and ethical challenges here, not to mention political ones. After a certain number, does it really matter, and why?

But what I do hope is very clear is that the period c. 1750-1914 was very violent.

Recognizing this violent period helps us understand better the murderous twentieth century, as I discuss in a bit more detail in the other essays. One observation is the contrast between the episodic convulsions of the twentieth century and the “smoother”, more consistent violence of the earlier period.

Before examining this database, however, users should be aware of the following issues by way of introduction:

There are obvious challenges about violence and its aftermath, such as displacement, famine, disease, etc. Remember that the data tries to incorporate all these sequalae.
I have not yet proofed the information, nor have I checked it against what is in the book. (Students, whatever you do, do not use this material for a research paper!) There may, for example, be numerous instances of overlap.
The list is quite incomplete. There are, for example, just a very few inclusions relating to the Russian Empire and the Ottoman Empire. These two empires were massive, and like all empires engaged in violence. The database also misses many other polities. To take just one example, it really does not reflect the rise and fall of the Comanche in North America, or many other instances of intra-native violence in North America.
The database is in general state/empire-centric. This is very important especially for the eighteenth and for much of the nineteenth centuries when the new means of destruction were in effect very democraticized. So, banditry and warlordism is weakly/unevenly represented in the database.
Dangers of overlap, as noted above.
I have added a very, very short and incomplete bibliography if you want to delve into some of this in more detail.
To conclude: our information is very, very uneven. Some estimates will be solid, others much less so. These figures should be taken as very rough approximations, but I hope they at least shed a little light on the past.

But I hope the point is clear. The period c. 1750-1900 was filled with violent conflict.

You can access the database here. Remember that I am more or less constantly revising, so totals may not align with above: 23MARCH-2026-Deaths_Outside_Core-W_-Europe_1750_1914_with-Empires (2)

PLEASE DO NOT REPRODUCE WITHOUT PRIOR EXPRESS PERMISSION

You can access a very, very short bibliography here: Further reading, Counting the Dead