From Glenn Loury's article in The Boston Review:
According to a 2005 report of the International Centre for Prison Studies in London, the United States—with five percent of the world's population—houses 25 percent of the world's inmates. Our incarceration rate (714 per 100,000 residents) is almost 40 percent greater than those of our nearest competitors (the Bahamas, Belarus, and Russia). Other industrial democracies, even those with significant crime problems of their own, are much less punitive: our incarceration rate is 6.2 times that of Canada, 7.8 times that of France, and 12.3 times that of Japan. We have a corrections sector that employs more Americans than the combined work forces of General Motors, Ford, and Wal-Mart, the three largest corporate employers in the country, and we are spending some $200 billion annually on law enforcement and corrections at all levels of government, a fourfold increase (in constant dollars) over the past quarter century.
Never before has a supposedly free country denied basic liberty to so many of its citizens. In December 2006, some 2.25 million persons were being held in the nearly 5,000 prisons and jails that are scattered across America's urban and rural landscapes. One third of inmates in state prisons are violent criminals, convicted of homicide, rape, or robbery. But the other two thirds consist mainly of property and drug offenders. Inmates are disproportionately drawn from the most disadvantaged parts of society. On average, state inmates have fewer than 11 years of schooling. They are also vastly disproportionately black and brown.
Italics mine. I find that to be an utterly astounding statistic. The size of our workforce dedicated to imprisoning mostly-non-violent Americans is not merely equal, but significantly larger, than the workforces of our country's most massive three employers combined. And the population they exist to serve keeps growing, even as crime keeps dropping. Why? Because that's how we want it:
One simple measure of punitiveness is the likelihood that a person who is arrested will be subsequently incarcerated. Between 1980 and 2001, there was no real change in the chances of being arrested in response to a complaint: the rate was just under 50 percent. But the likelihood that an arrest would result in imprisonment more than doubled, from 13 to 28 percent. And because the amount of time served and the rate of prison admission both increased, the incarceration rate for violent crime almost tripled, despite the decline in the level of violence. The incarceration rate for nonviolent and drug offenses increased at an even faster pace: between 1980 and 1997 the number of people incarcerated for nonviolent offenses tripled, and the number of people incarcerated for drug offenses increased by a factor of 11. Indeed, the criminal-justice researcher Alfred Blumstein has argued that none of the growth in incarceration between 1980 and 1996 can be attributed to more crime.
This growth in punitiveness was accompanied by a shift in thinking about the basic purpose of criminal justice. In the 1970s, the sociologist David Garland argues, the corrections system was commonly seen as a way to prepare offenders to rejoin society. Since then, the focus has shifted from rehabilitation to punishment and stayed there. Felons are no longer persons to be supported, but risks to be dealt with. And the way to deal with the risks is to keep them locked up.
This, of course, affected policing priorities, to the point that researchers now find that "higher incarceration in a given neighborhood in one year seemed to predict higher crime rates in that same neighborhood one year later," as policing becomes more intensive and less-lenient in high crime neighborhoods, parolees returning to their homes are closely monitored, and sentencing laws for repeat felons come into play. In other words, imprisonment begets imprisonment -- because we want these populations imprisoned.
