Bryan Caplan  

U.S. Executions Per Capita Have Been Falling for 400 Years

PRINT
Vipul Naik's Analysis of the O... Is Fiat Money a Bubble? ...
The death penalty has fallen into disfavor in recent years.  But what's the long-run trend?  I've intermittently wondered about this for over a decade.  Last night, I finally decided to check.

I found a time series of total U.S. executions from 1608-2002 here, broken down (roughly) by quarter-century.  I found a time series for long-run population here.  The numbers are irregular during the early centuries, with decennial censuses since 1790. 

Since the periods for the two series don't perfectly match, I just use the population numbers closest to the final year of each quarter-century of executions.  For example, 1650-1674 had 49 executions, and the year closest to 1674 for which population data is available is 1670 (when population was 111,900). 

Here's what I found:

deathpen3.jpg
The decline in executions per capita is perfectly monotonic.  Since executions are aggregated over quarter centuries, the pattern isn't utterly amazing, but I was still shocked by its consistency.  In historical perspective, capital punishment virtually disappeared over a century ago, when its incidence approached one-in-a-million per year.  "Ideas have consequences," you say?  Perhaps, but executions were in rapid decline long before liberal opponents of the death penalty began their crusade.

I can easily imagine my graph in a Julian Simon or Steven Pinker chapter on human progress and the decline in violence.  Even though I have no philosophical objection to the death penalty, it's hard not to interpret this 400-year pattern as a strong sign of human betterment.  Anyone care to say nay?



Comments and Sharing





COMMENTS (24 to date)
Doug writes:

Singapore averages 18 times the execution rate of the United States, despite having about 10 times less capital crime.

As anyone who has been to the city-state can attest a person feels safe everywhere. It's a serious quality of life issue that those in the West are barely even aware of. There's scarcely a square block in any American city where it's advisable for a young lady to walk alone at late at night.

These have serious costs. Would Singapore achievable level of crimes be possible without Singapore style capital punishment? Maybe, but the proof is in the pudding.

Should 10 million people in NYC feel afraid at night so we can save hanging one or two career criminals?

NZ writes:

Other factors I can think of besides people becoming nicer over time:

-changing demographics
-imprisoning lots of people James Q. Wilson style

@Doug: I would think demographics and levels of racial diversity are probably the main factors that determine how safe a given city feels at night.

arne.b writes:

If there had been a similar trend in the number of slaves per capita up to 1860, should that have been interpreted by someone with "philosophical objections to" slavery "as a strong sign of human betterment"?

Filip writes:

I'm not sure per capita numbers are useful in this case. Some thoughts here.

[shortened url expanded to full url--Econlib Ed.]

Shane L writes:

Interesting stuff!

Countries probably couldn't afford to imprison lots of people centuries ago. See Britain instead shipping its convicts to Australia, or forced to use horrible rotting old ship as floating prisons.

Another aspect is probably the real long-term fall in crime in developing countries.

And finally a much older Christian or Enlightenment tradition of opposition to more severe forms of punishment like torture, corporal punishment and execution. So a mixture of ideas and economic growth, perhaps.

Bob Knaus writes:

@Doug: This list of the 21 countries using capital punishment in 2011 contains no others as safe as Singapore.


You might consider that culture determines the use of capital punishment, rather than vice versa.

philemon writes:

@Doug

How did you arrived a the 18x figure? (I'm genuinely curious). According to the data linked from this, Singapore sentenced 26 to death between 2007-2011,and executed 4 between 2007 and 2010. The numbers for the US is 220 and 504. The population of Singapore was between 4.5 and 5 million in that time period, and that of the US, between 300 and 310 million.

There could be a parallel movement in numbers in Singapore over time as well. But I'll need to dig up the statistics, assuming they are available.

I'm not totally convinced that the use of the death penalty is a good explanation if cited in isolation from other factors (culture, general efficiency of the police and government, economic conditions, etc.).

By the way, I do concur that a person feels safe (almost) everywhere in Singapore. Though it is possible to enjoy comparable safety in many specific neighborhoods in US cities, there are also many many places where one does not feel safe (e.g., where it is NOT advisable for a young lady to walk alone at late at night, to use Doug's metric). Such places are very rare to almost nonexistent in Singapore, where the main public education message of the police force is "Low Crime Does Not Mean No Crime" (i.e., warning against complacency).

Brandon Berg writes:

This list of the 21 countries using capital punishment in 2011 contains no others as safe as Singapore.

Japan's not on the list because they didn't execute anyone in 2011, but they did execute seven people this year.

Bostonian writes:

Part of the reason for fewer executions is that the judicial process has become absurdly drawn out. Nidal Malik Hasan killed 13 soldiers on November 5, 2009. Almost two years later, his trial has not started yet. There is currently a dispute over his beard. This is not progress.

RPLong writes:

I would like to see how that graph changes when you factor into account foreign nationals killed in undeclared wars and people killed during police skirmishes before they ever had the chance to be brought to trial.

Finch writes:

> If there had been a similar trend in the number
> of slaves per capita up to 1860, should that have
> been interpreted by someone with "philosophical
> objections to" slavery "as a strong sign of human
> betterment"?

Yes, of course it would. Suppose we'd abolished slavery because economic progress led to it making more sense to pay people for increased productivity, rather than through a war. Would that have been a bad outcome?

Similarly, we may be using capital punishment less at least partially because we are committing fewer crimes that deserve it, rather than because of some philosophical crusade. I'd say that's just fine.

Ken B writes:

Interesting point about the decline predating its becoming a political issue. That suggests that perhaps it became a trendy issue because it was declining. Another case of signalling in other words.

John Voorheis writes:

"Yes, of course it would. Suppose we'd abolished slavery because economic progress led to it making more sense to pay people for increased productivity, rather than through a war. Would that have been a bad outcome?"

I would call a couple of million slaves waiting in bondage for the market to discipline southern slaveholders rather a bad outcome.

egd writes:
I would call a couple of million slaves waiting in bondage for the market to discipline southern slaveholders rather a bad outcome.
Others might call a couple of million war deaths rather a bad outcome.

If we could abolish the death penalty today by killing 100,000 volunteers, would it be a good exchange?

I would think demographics and levels of racial diversity are probably the main factors that determine how safe a given city feels at night.

I'd be very surprised if that 'racial diversity' turned out to be true. Most violent crime is between people of the same race.

Thomas Sewell writes:

I'd be much more interested if you put this graph up with a corresponding graph on death penalty eligible crime rate changes over time.

As executions tend to be the result of specific crimes, using a per capita measurement without reference to the crime rate for capital crimes doesn't really tell us if it's just the crime rate dropping, or if it's less people being executed because the system is taking longer, or because more people don't get a capital sentence, or what.

Finch writes:

> I would call a couple of million slaves waiting
> in bondage for the market to discipline southern
> slaveholders rather a bad outcome.

My premise was that we achieved victory against slavery by some means other than convincing everybody it was evil. I wasn't endorsing slavery; I don't know how you could reasonably have read my post that way.

In the case of capital punishment, the philosophical argument has mostly failed. But we have small numbers of executions. This is a good thing. If the number was zero because the sort of crime that warrants execution never happened, I'd be very happy, and I wouldn't spare a thought for the failed debate.

bastiat writes:

@Doug, how can one possibly feel "safe" when Big Brother is watching you? Are there *really* classical liberals out there putting the arguments for a police state?

Evan writes:

@NZ

@Doug: I would think demographics and levels of racial diversity are probably the main factors that determine how safe a given city feels at night.

Glancing at Singapore's demographics, it looks like it has a similar level of racial diversity to the USA. In particular both countries have one race that makes up about 75% of the population, another race that makes up about 13%, and some other races that make up smaller percentages. Of course, what specific races those are different in each country, which might be important.

In fact, Singapore might actually more diverse than it appears at first glance, because I think the statistics I was looking at might not count guest workers, who make up a sizable percentage of the population.

So I don't think ethnic homogenity by itself is a valid hypothesis to explain why Singapore feels so safer than the USA, simply because Singapore isn't significantly more ethnically homogenous than the USA.

Mike Rulle writes:

If we just stick to the data and not hypothesize its cause (without looking at further data) there is nothing we can say about whether we have improved or not. For example, it could be just a function of the same crimes being committed as before but less capital punishment being implemented as punishment. If less capital punishment per se is seen as a good of itself (which it might be) then on that score I could agree.

I do not know if we have less crime per capita. We just have less executions per capita. Faststats says we had 17000 homicides in the last year they reported. If all those who committed homicides were executed, our rate would be back to the 1780s. My impression is we executed people for lesser crimes than that in the 1700s.

My belief (and that is all it is) is we have less crime and that old data is very bad by understating actual violence----however, my only point is this graph is not conclusive evidence of such.

zman writes:

I will say nay because that chart is overwhelmed by two other issues that are very negative: 1) the huge increase in imprisonment. 2) the, relative to other points in the 20th century (and especially given how many people we put in prison), high crime rate. Those two things are both catastrophes for millions of families. And given how utterly terrible our prisons are, I am not at all convinced that our justice system was not more humane 80 or 90 years ago when there were quick trials of heinous crimes followed by swift justice. I do not believe 100 years from now people will believe us to have been more liberal than our ancestors in terms of punishing crime.

James A. Donald writes:

The number of habitual criminals, the number of people who wake up every morning intending to burglarize or mug someone, who every afternoon check out the neighborhood for a house to burglarize, is much higher than the number of people who get the death penalty.

This is a major and massive deterioration in the quality of life.

The pareto optimal solution to crime

Fines for all crimes if the criminal can pay. The victim is compensated sufficiently to go a fair way to making him whole, but no so much as to unduly encourage false accusations. Everyone involved in catching the criminal is also paid, an amount whose total should be be roughly equal to damage caused by the crime divided by the probability of getting caught. Total assessed compensation for any category of crime should approximately equal the damage caused by that category of crime . If the criminal cannot pay, which he usually cannot, is sold into chattel slavery. If no one wants to buy him, or if his new owner, quite possibly the victim, feels like executing him, the criminal is executed. If he runs away and is caught, executed.

If his price is bid up to the value of the fine, he gets to choose who buys him (probably friends or relatives bailing him out)

Suppose someone shoplifts something of small value, but cannot pay the small compensation. Well then, we would expect his friends and relatives to bail him out.

What? He has no friends, and his relatives disown him? Obviously, he will not be missed.

This solution is optimally economically. To the extent that we deviate from it, it is economically inefficient, inefficient in ways that impose a cost on me and people like me.

Ely Spears writes:

So in addition to becoming a pirate, one whiz bang way to reduce global warming is to increase per capita executions in the US!

hana writes:

The 400 year narrative creates quite a nice hockey stick, but I don't believe the change reflects anything other than what Mike Rulle comments above. If it is broken down by time period and what were the capital crimes, it would be much more informative.

Assuming a relatively small, isolated, military type force at the beginning of the time line, I would expect executions might be a reflection of both mores and military discipline. In a small group this would skew the capital punishments per population.

During the next time frame there were a number of executions for 'witch craft'. The explanation for these would reflect mores again plus the influence of religious discipline inside a closed society. In one sense this would not be so much different than the military discipline. Again in a relatively small population this can skew the capital punishments per population.

A next time frame would include crimes that are no longer considered capital punishment through modern eyes. Stealing a horse from someone in 1850 in New England has much smaller ramifications that stealing the same horse in Nevada in 1850. In New England the ramifications would include the hindrance of a loss of capital or mobility, in Nevada at the same time the theft of horse might have meant a death sentence for the victim. So in looking at the capital crime rates, the relative peril of the victims would change the nature of capital crimes. In today's world the theft of a car is generally an inconvenience, one could imagine extraordinary circumstances that would make it a much more serious situation, but those are by definition extremely unusual. These type of perilous incidents occasionally turn up as capital crimes.

I would suggest that as MR mentions above, the change in executions may be strongly attached to the re-categorization of capital crimes, not any improvement in society.

Comments for this entry have been closed
Return to top