The way justice decides to actually imprison convicted people varies across countries, and consequently, prison rates (number of prisoners per 100 000 inhabitants) are very disparate. In the top 10 countries, you can find countries as different as the United States of America (665 prisoners/100 000), El Salvador (618), Thailand (558), Turkmenistan (552), Cuba(510) or the Maldives (499) and it seems difficult to identify a pattern. It might depend on justice structure and independence, type of political regime (democracy, autocracy, dictature…), the ability the transform imprisonment into another type of sentence as well as global criminality level.
Comparing with a criminality proxy (homicide rate), it appears that prison rate is definitively not correlated to the level of delinquency. Some countries face a high level of homicide and have a consistent level of incarceration (El Salvador, Kenya, Brazil….). At the other end of the spectrum, Iceland, Japan and Burkina Faso have a low rate of criminality and a limited rate of imprisonment. Some other countries (mainly from sub-Sahara) have an interesting profile with a high level of criminality and a low rate of people in jails. In this situation we might question the ability of these states to enforce laws. The last remarkable group of countries is composed of the USA, Cuba, Turkmenistan, Thailand and Guam. Indeed, they have homicide rates close to average but an extreme prison rate.
Homicides can’t be considered as a comprehensive proxy of all types of offenses and the graph could be completed with other areas of delinquency that would explain large differences (for example, India and Thailand), but one could wonder if imprisonment is an appropriate answer to decrease elinquency and criminality. To perform a more in-depth study, we should first control all variables using socio-economics environment and crimes typology. Then we could try and assess the complex causal link(s) (as defined by Granger) between imprisonment and criminality.