Here’s a summary of the U.S. Department of Justice National Institute of Justice’s “Five Things about Deterrence” July 2014:

  1. The certainty of being caught is a vastly more powerful deterrent than the punishment.
  2. Sending an offender to prison isn’t a very effective way to deter crime.
  3. Police deter crime by increasing the perception that criminals will be caught and punished.
  4. Increasing the severity of punishment does little to deter crime.
  5. There is no proof that the death penalty deters criminals.

DOJ Source: Daniel Nagin, “Deterrence in the 21st Century,” in Crime and Justice in America: 1975-2025 (ed. Michael Tonry, University of Chicago Press, 2013).

Bottom line: more police, less prison time. The DOJ piece says prison doesn’t have much deterrence effect – but if it had no deterrence effect, being caught would have no deterrence effect. How much prison or jail time is enough to make one think twice before committing to some criminal act? I’m thinking one or two years on average, maybe five years max for non-violent crimes. These sentences would be more in line with incarceration policy in other developed countries (see, for instance, Finding Direction: Expanding Criminal Justice Options by Considering Policies of Other Nations by the Justice Policy institute).

Since lengthy sentences have little deterrence value, why should prison time for non-violent crime ever be more than a few years? Let’s reduce prison sentences for non-violent offenders and put the money saved to what truly has deterrence value: more police. And then couple greater police presence with more effort to re-integrate ex-cons into society. Reinstate eligibility for school financial aid. Expand tax credits for employers hiring ex-cons. Increase job placement services and other forms of integration assistance after release.

We should also let out many of the aging prisoners who pose little threat to public safety, even though they originally committed violent crimes. Violence is mostly a young man’s game. Just 14% of violent felons are 40 or older. Recidivism rates also decline with age and, ironically, those with the lowest rearrest rates are those who had been in prison for homicide.

If we let out a bunch of old codgers and reduced prison sentences for most non-violent offenders (about half of the state and federal prison population) so that the overall prison incarceration rate goes down by a third and could then transfer the savings to police salaries, how many extra police would that pay for? About 200,000 – increasing US police numbers from 780,000 to 980,000.

Think about how 25% more police would boost crime solve rates. Right now these rates are abysmal: in 2013 the US crime clearance rate in the United States for all crimes was 40.6% with the following breakdown: 14% for motor vehicle theft, 13% for burglary, 21% for arson, 22% for larceny, 29% for robbery, 40% for forcible rape, 58% for aggravated assault, and 64% for murder and non-negligent homicide.

Of course, a higher solve rate would initially increase the jail and prison populations – but in pretty quick order – thanks to the robust deterrence effect of increased police presence – crime rates would go down, as would the number of incarcerated.

How I arrived at my figures:

Per the Bureau of Labor Statistics: the median annual pay for a correctional officer is about $40,000. In 2012, there were about 470,000 correctional officers in the US. The median annual pay for police officers is about $57,000. In 2012, there were about 780,000 police officers in the US. It costs an average of $31,000 year per inmate incarcerated (ranging from $15,000-$60,000, depending on the state). There are about 1.5 million federal and state prisoners. (I’m leaving out jail inmates for this discussion- it’s complicated enough, thank you).

The above figures would work out to a current state and federal prison budget being about $45 billion. However, a Center for Economic and Policy Research report (The High Budgetary Cost of Incarceration by John Schmitt, Kris Warner, and Sarika Gupta; June 2010) say it’s about $52 billion, so I’ll go with the higher figure.

Translating reductions in sentence length to reductions in overall prison population to reductions in prison budgets is not a straightforward process, so I’m again going to rely on the CEPR Report. According to this report, lengthy prison sentences are largely responsible for the currently high US incarceration rates and if incarceration rates went down to 1993 levels, there would be savings of almost a quarter of total correction budgets. That means about $12 billion in annual savings that could be used to pay for more police.

How many police would that pay for? About 200,000, increasing police numbers from 780,000 to 980,000.