Prisons are usually newsworthy only when their inmates riot, but of late they have become the focus of more general, and on the whole more constructive, attention. The revival of interest in the deterrent effects of the criminal-justice system, the spreading “prisoner-rights” movement, the decisions of various federal judges to close (or to threaten to close) various overcrowded or indecent jails and prisons, the sudden increase in the size of the prison population, and the emerging debate over whether new prison construction is desirable have all combined to thrust a venerable and unpleasant institution onto center stage.
The debate over prisons raises many issues—the length of sentences, the function of parole, the level of amenity, the feasibility of rehabilitation, the rights of inmates—but central to most of these issues are some mundane but rarely examined constraints. If one wishes to award longer sentences to those who now go to prison, or short sentences to a larger fraction of those convicted of a crime, one must find a place to put them. If one wishes to increase the level of amenity, it presumably requires lowering the density of prison populations: it is hard to imagine how to make any institution less brutal if three or four persons are assigned to a cell designed for one, or if persons must sleep in the corridor or on the floor. Among the rights that most inmates certainly cherish is the right not to be abused by fellow prisoners, yet that can scarcely be insured without producing a greater degree of privacy, and accordingly supplying more square feet, for each inmate. In short, the sheer capacity of the prison system will importantly influence what, if anything, can be done toward achieving various objectives.
In general, we do not know what the “capacity” of our prison system is. There are about 400 prisons, 158 community centers and work farms, and 33 other (reception, diagnostic, medical, or classification) facilities operated by the states. Half of all of these are located in the South. These figures, the best available, were compiled by the Census Bureau in January 1974, but that agency admits that there may be more or different institutions—it is hard to be certain what is a distinct facility, some facilities have more than one function, and half of the institutions in one state (Massachusetts) did not even bother to reply to the questionnaire. The survey did not indicate what the “capacity” of each facility was and no survey probably could, inasmuch as many correctional administrators have long since had to put more inmates into a prison than it was designed to receive and in consequence have administratively redefined what constitutes its capacity.
We do know that the great majority of prisoners are in conventional closed prisons, not in work farms or community centers. Though the movement to place more convicted offenders in these “open” facilities has grown of late, less than a tenth of the correctional population is in community centers and less than a seventh on prison farms. Though most inmates are in real prisons, the popular image of these facilities as uniformly ancient and decrepit is not necessarily true. Nearly half of all prisons were built since 1949 and three-fourths were constructed since 1924. There are at least 50 prisons, however, that were built in the 19th century, several around the time of the Civil War.
If we cannot measure the capacity of prisons, we can at least count their populations. Except for the period during and just after World War II, that population has been remarkably stable. There were 180,000 persons in state and federal prisons in 1939, and there were still about that many in prison fifteen years later, in 1954. During the 1950’s, a period of relatively stable or only moderately rising crime rates, the prison population went up sharply—from 166,000 in 1950 to 213,000 in 1960. Then, during the 1960’s, a period of rapidly increasing crime rates, the prison population actually declined from 213,000 in 1960 to 196,000 in 1970.
Why the prison population should go up when crime is not serious and go down when it has become a national crisis still defies explanation. One possibility is that certain judges came to believe that prison was an inappropriate way to treat criminals—because these facilities did not rehabilitate, or because they were unpleasant places to live, or because prison was generally “illiberal.” This explanation is lent plausibility by the fact that many of the largest reductions in state-prison populations during this decade occurred in states, such as New York, Michigan, Pennsylvania, Minnesota, and Hawaii, with a strong tradition of liberal politics and presumably a liberal judiciary. Yet declines also occurred in Mississippi, West Virginia, Georgia, Alabama, Kansas, and Kentucky, hardly hotbeds of leftists.
One of the most intriguing theories devised to explain the relative stability over long periods of time in imprisonment rates is that advanced by Alfred Blumstein, Jacqueline Cohen, and Daniel Nagin of Carnegie-Mellon University. They show that, not only in the United States but in Norway and Canada as well, the imprisonment rate—the number of persons in prison per 100,000 population—is remarkably stable. Between 1930 and 1970 in this country it averaged around 110 with only small variations around this mean (the standard deviation was less than 9). In the late 1930’s, the rate went up to abnormally high levels, in the 1940’s down to abnormally low ones, in the 1950’s it increased, and in the 1960’s it decreased. Their theory is that society penalizes less serious offenses at times when serious crime is not a problem and lets off minor offenders when serious crime has become endemic. If murder is relatively uncommon or at least declining in rate (as it was during the 1950’s), society will worry more about shoplifting, prostitution, and auto theft. When murder, assault, and robbery are increasing rapidly (as they did in the 1960’s), society will ignore the shoplifter and seek—de facto if not de jure—to decriminalize prostitution.
The theory is quite suggestive, but thus far unproved. Among other difficulties, it does not account for the substantial variation among states in the use of prison. Crime rates during the 1960’s were soaring in both Michigan and New Jersey, but whereas the former state reduced its prison population, the latter increased it despite similar rates of population change. Massachusetts officials have long been preoccupied with the dangers of overcrowded prisons and perhaps in consequence of this kept their prison population constant during the 1960’s despite a tripling of the serious crime rate. South Carolina, on the other hand, allowed its prison population to rise by more than 30 per cent even though it meant increased over-crowding. For some states, the capacity of the prison system is apparently an important constraint on the sentencing practices of judges, but in other states it is not.
In recent years, the prison population has started to rise again. In early 1976, according to Corrections magazine, there were nearly 250,000 persons in prison, the largest number since figures began to be kept. In the Blumstein, et al., theory, this was to be expected—after allowing the imprisonment rate to fall during the 1960’s (from 119 per 100,000 in 1960 to 97 per 100,000 in 1970), it was inevitable that pressure to respond to the crime wave would drive it back up to its traditional average of around 110. If that rate is applied to the estimated 1975 total population, it would produce a prison population of 234,000, significantly below the actual figure.
The reasons for the increase above expected levels are not hard to find. The population today is more youthful than at any other time in this century, and since young people are disproportionately criminals, it is not surprising that prisoners are growing in numbers faster than society generally. Furthermore, many of the young offenders who were first convicted in the 1960’s are now being convicted for the third or fourth time and as a result can no longer expect to avoid prison through probation or suspended sentences. In addition it is possible that judges are becoming more severe, but since we have no national statistics on sentencing that can only be a conjecture.
There is nothing magical about the long-term stability of the imprisonment rate. Society can choose a higher or a lower rate, in part by altering its prison capacity. That can be accomplished by either building more facilities, reducing the sentences of those going to prison, or changing the mix of persons in prison. The first policy results in an absolute increase in prison size, the second permits more persons to be sentenced to prisons despite a constant capacity, and the third permits one to find prison space for increased numbers of more serious offenders.
No accurate national figures exist on changes in total prison capacity. The Federal Bureau of Prisons, which has in confinement about 10 per cent of all prisoners, has, in the last five years, opened thirteen new facilities with a total capacity of over 4,000 inmates but still feels that its system is overcrowded by about 5,000 inmates. Among the states, major new building programs are under way or proposed in Florida, Alabama, North Carolina, Ohio, and elsewhere. The National Moratorium on Prison Construction, a lobbying group sponsored by the Unitarian-Universalist Service Association and the National Council on Crime and Delinquency, which opposes prison construction, reports that it has heard of over 500 planned institutions. Whatever the actual figure, it has so far been inadequate to alleviate the severe over-crowding that now exists and that will, almost certainly, become worse.
Because capacity has not kept up with inmates, there has occurred a dramatic shift in the kinds of crimes for which persons go to prison. In 1960, the federal Bureau of Prisons counted about 151,000 persons in state prisons. In 1974, the Census Bureau (which has taken over the job of keeping track of prison populations) estimated, on the basis of a survey of about 10,000 inmates, that there were 188,000 persons in state prisons. The increase in size is partly an artifact of who was counted; in 1960, some states gave incomplete or inaccurate figures and persons not in closed prisons were often omitted. In 1974, the coverage was broader and included persons in certain state mental hospitals as well as those on work-release programs and in community centers. Because the figures are not entirely reliable, only gross differences should be noted. The striking finding is that prison is increasingly reserved for only the most serious offenders. Persons convicted of homicide, robbery, and assault made up one-third of the state-prison population in 1960 but nearly half in 1974. There are nearly twice as many murderers in prison today as in 1960 and 70 per cent more robbers. By contrast, the number of burglars, thieves, and auto thieves in prison has actually declined—from about 55,300 in 1960 to about 49,200 in 1974. (During this period, the reported rate of these crimes increased more than fourfold.)
Some have argued that we do not need more prisons because the ones we have are crowded with persons serving sentences for various “victimless” crimes. This was not true in 1960 and it is not true today. Only 10 per cent of the 1974 inmates were serving sentences for drug offenses and only one per cent for sex offenses other than forcible rape. Of the drug offenders, nearly half were serving time for trafficking in hard drugs, such as heroin. Thus, only 6 per cent of the inmates were doing time for merely possessing a drug or trafficking in marijuana. The number of persons in prison for gambling is not reported, but could not exceed one per cent.
Others have argued that prison should be reserved for repeat offenders and not used for persons convicted for the first time. By and large, that is already the case. Less than one per cent of the inmates had never been sentenced before; 28 per cent had served four or more prior sentences.
Prison inmates are, as one would expect, disproportionately drawn from among the poorly educated and those with low income. About half are black (in 1960 only a third were) and most never finished high school. Contrary to what one might suppose, inmates are not typically unemployed during the month preceding their arrest. Two-thirds were employed, and almost all of these said they were employed full time. Of those unemployed, the majority said they had not been looking for work and did not want work.
The other way society can alter its prison capacity is by changing the amount of time the average inmate serves. The shorter that period, the more space each year for additional inmates. Unfortunately, no good figures are available on time served by inmates of all state prisons. We do have such data for certain jurisdictions, however, and we can make an estimate of it for prisons generally. Overall, the length of time served in prison seems to have remained pretty much the same over the last decades or so: for the average inmate, about two years (he may be sentenced to much longer terms, but offenders are actually incarcerated for no more than one-third to one-half of their nominal sentences). This would suggest that nothing has changed. But that is a mistaken inference. Recall that the average prisoner is more likely today to be a murderer or robber and much less likely to be a burglar or auto thief. If time served has remained constant, it can only mean the time served in prison for more serious offenses, such as murder and robbery, has become less.
It is not clear, of course, whether any changes in time served were the result of judicial decisions. Other agencies, such as parole boards, can and do decide how long a person will stay in prison whatever the initial sentence. How this has changed is vividly shown by the experience of the federal correctional system. Federal judges, like their state counterparts, sent fewer persons to prison in 1970 than in 1960-30 per cent fewer, as it turns out—despite the rising crime rate. For those they did imprison, however, they issued longer sentences: in 1960, average sentence length was about 27 months, but in 1970 it was over 38 months. At the same time, however, the proportion of the sentence actually served fell—from about 64 per cent in 1960 to 51 per cent in 1970. As a result, the actual time served remained about constant (around a year and a half) despite the fact that federal prisons in 1970 had a larger proportion of inmates who had committed more serious crimes, such as robbery, assault, and firearms-law violations.
In sum, we have adapted to higher crime rates by reducing the proportions of arrested persons going to prison (a pattern that recently may have been reversed), changing the mix of prison populations to emphasize serious over less serious offenders, keeping more or less constant the average amount of time served despite the fact that those in prison are guilty of more serious crimes, and allowing overcrowding to occur in many systems (especially in the South).
From many different, and even competing, perspectives, this has not been a rational response. If you believe that increasing the percentage of convicted offenders who are imprisoned will deter others from committing crimes, then a declining prison population and a declining imprisonment rate are mistakes. If you believe that more serious offenders should be kept off the streets longer, then a constant sentence length at a time when the prison population is coming disproportionately to consist of serious offenders is a mistake. If you believe that prison conditions are inhuman and indecent, then allowing overcrowding to occur in many states makes difficult or impossible the provision of higher levels of privacy and amenity. If you believe that “white-collar” criminals should be given prison sentences more frequently (either to deter others, to exact retribution, or to prevent the burden of prison from falling overwhelmingly on the poor and the black), then maintaining a limited prison capacity is a mistake, because white-collar (and other nonviolent) offenders will inevitably be displaced by blue-collar, violent ones. (Between 1960 and 1970, the number of persons in federal prison who were convicted of income-tax violations, embezzlement, fraud, and forgery declined while the number convicted of robbery increased.)
In short, increasing the capacity of our prison system is consistent with, if not indeed mandated by, a variety of policy goals that are often thought to be in opposition. It is not clear, of course, that merely having more “prisons”—and by that I mean all variety of facilities, community-based as well as maximum-security—will induce judges to use them more, parole boards to leave serious offenders in longer, or wardens to devote more effort to producing decent and humane environments. It would seem, however, that for the nation as a whole, if not for every state, the more or less fixed capacity of our prison system has been allowed to become a major constraint on, or excuse for, sentencing policy. William A. Shaffer of MIT has shown, through a careful analysis of the Massachusetts criminal-justice system, how this constraint operates in practice. In that state, he suggests, the shortage of prison capacity is the critical variable affecting the ability of the system to lower crime rates—much more so than police resources or court decisions.
The cities and states, of course, are often pinched for money, in some cases led by officials who believe that only harsh correctional systems are of any value, and usually caught in the middle between citizen groups that want less crime and those that want to block the construction of new prisons. The problem will become more acute as prison populations continue to rise, as they have begun to do. I suggest that the next national administration, whether motivated principally by a concern for victims or inmates, has an obligation to examine the federal relationship to the state-prison system and to think seriously about the desirability of supplying both more funds for expansion of that system and higher standards for its operation.
1974—Estimates derived from National Criminal Justice Information and Statistical Services, Law Enforcement Assistance Administration, “Survey of Inmates of State Correctional Facilities, 1974,” National Prisoner Statistics (SD-NPS-SR-2), Table 4. Because 1974 figures are estimates from a sample survey, they have been rounded off to the nearest hundred.