Data covering prisoners world-wide, collected through the pioneering efforts of Roy Walmsley and carried on through the University of Essex’s International Centre for Prison Studies, are among the best quality international statistics available. Prisoner statistics are a simple, intuitive population statistics. Moreover, state administrative institutions are more closely connected to prisoners than to the resident population in general. Compared to international statistics on the labor force or economic output (GDP), prisoner statistics are conceptually clearer and more readily amenable to reliable collection.

Compiling comparable prisoner statistics world-wide is a major challenge. A prison census on a particular date can produce a different figure from a year-long average of the number of prisoners present. Imprisonment prevalence statistics, often confusingly described as “imprisonment rates”, differ from imprisonment incidence statistics. Distinctions between juvenile delinquents and prisoners vary. Persons are put in prison for different reasons and terms, and only a subset of all prisoners might be counted. Some types of punishment, such as part-time confinement and work camps, raise questions about appropriate counting rules; and, in a small number of instances, the sex of a prisoner is not merely a matter of common sense.

Intimations of statistical problems can intimidate persons not experienced with empirical work. Focusing on anecdotes, literary representations, and detailed local studies doesn’t provide an escape from imperfections in evidence and the need for scholarly caution. Despite its imperfections, available aggregate statistical data on men and women in prison in jurisdictions around the world are useful for understanding the common human experience of imprisonment.

One regularly finds in the scholarly literature warnings that statistics are imperfect and should be handled cautiously. Those warnings apply equally well to any scholarly work and to life in general. One should not fear embracing statistical reality with reason and acting based on information.