Book Review: Death by Government

Death by Government
R. J. Rummel
Website
Blog
Transaction Publishers (March 1, 1997)


Power tends to corrupt, and absolute power corrupts absolutely
Lord Acton

Power Kills, absolute power kills absolutely. It is advised to get the readers attention in first sentence. Professor Rummel takes that advice to heart in the opening sentence, following on Lord Acton’s famous quote. This book is a census of the people who have been murdered by governments and why. He explains his thesis in the first chapter and the rest of the book is a review of the governmental deka-murders mega-murders, i.e. ten or one millions.

This not limited to genocide. Genocide is the destruction of a nationality, people religious group, most likely by murder, but deportation or dispersal would meet the definition. He uses the terms politocide to describe a government killing for political reasons and invents the term democide, to describe the any situation where a government murders. The 20th century genocidal murders are less than half of all democide. He excludes persons killed by lawful military action or executed for what are commonly considered capital crime. The first function of a government is to protect it citizens/residents from violence, democide is an abomination to the basic purpose of a government.

At this point we should stop and define the political science meaning of “power”. (Rummel assumes that anyone who picks up a book like this would have taken Poly Sci 101 and gives the background minimal treatment) As a neutral term “power” is the ability to influence or cause or do something. The something may be good bad or indifferent. One of odd things one sees is an activist who is unsuccessfully attempting to achieve his goal accusing the opposition of being power hungry. Actually the activist wants the ability (power) to accomplish his goal, he is down right hungry for power. His opposition most likely has enough ability (power) to frustrate the activists campaign without much effort, they are not power hungry they have it. Power should be distinguished from Authority – the right to exercise power, Responsibility - the duty (when where how) to exercise (or not) power, and Accountability – the ability of others to reward or punish for the exercise of power. The last three are part of the structure to control the use of power and prevent abuses.

The main thesis is that Governments or their agents will murder whenever they have absolute power, but when there is a dispersal of power which is accountable checked and balanced (i.e. a “democracy”) this is much less likely to happen. Democracy is in quotes because while he considers a modern liberal democracy preferable he will allow a larger definition. This would include any society where the effective power is distributed among the middle and lower strata of society as well as the top levels. It is the distribution of actual power that counts, some of the deadliest governments this century had externally democratic constitutions but actual power was in the absolute control of a small minority, conversely some of the monarchs of pre-modern times were in theory absolute rulers but in fact had little power without a consensus of at least the middle strata of society. Power can be good in the sense that if even the least members of society have the power to control their their own lives. The essential point is the wide distribution of power. And it is just not a national level, if the agents of even a democracy are in a localized place with absolute power it can be abused. For example an unsupervised night shift in Abu Ghrab prison.

When power is dispersed widely, horizontally and vertically, for a government to accomplish something requires the buy-in of a large percentage of the population. This can be unpopular with those who desire to make radical changes since it can difficult to build a concensus. Where power is concentrated there is little check on those who hold power, they can accomplish something quickly, but if they decide to murder a portion of the population they can’t be stopped. Rummel considers it a virtual certainty that sooner or later this will happen. The slowness and uncertainty of action caused by a wide dispersal of power is preferable to a certain democide when opower is absolute.

At first one is inclined to read the book and say “see my enemies are really bad guys”, but after reading a little more you see your own government or political movement, either in fact or potentially. The deadening recital of demicdes demonstrates that any nation movement or group can turn in to killers if given enough unchecked power. The strongest objections I found to his thesis on a Google search came form people who are part of movements that are trying to aquire enough power to accomplish radical change. Not doubting their good intentions, if Rummel’s theory is correct, when they achieve the power to accomplish their goals they will be added to the list of democide killers. Clearly now the burden of proof is on any group wanting extensive power to prove that they can’t abuse it.

Reviewing Rummels data it seems the 20th Century has had more per capita democide than previous centuries. The modern bureaucratic state can do many things more efficiently than it’s predecessors, including democide.




I would have liked to see him make a distinction between democide by murder and manslaughter. It seems that some of his examples were cases of the government attempting to do something in negligent manner causing massive deaths rather than intentionally killing. The distinction seems important to me to remind governments of the massive unintended secondary and tertiary effects a governmental action can have. This is a reason why the dispersal of power is important, the people who will be effected will be able oppose the action. It seems me that even with this distinction, there was per capitia more murder democide in the 20th Century than in previous centuries.

Rummel points out, that the numbers of killed are estimates. Some of his critics question his methodology. On his website, see above, he has a good explanations of his methodology and why he adopted them. His estimates are higher than the usual figures I have seen. For pre-modern times he accepts the figures in the original sources: data which is often hyperbole rather then mathematical, but the democide happened. One could quibble forever I suppose, but really even if he is 50% over, the value of the book is in the description of the process that produces democide.

I think that a specialist in a given country may have some corrections of the “it was Jones not Smith” category, an occupational hazard of all latitudinal studies, but the value of this study comes from the cross study of many democides so we can see the common process.




This is an important book, a well written book, a must read book, especially for any one interested in politics government military or social justice, but somewhere around 100 million dead I just could not call it a good book.