How much does imprisonment cost a state’s taxpayers? The question is conventionally answered simply by looking at the budget of the state’s department of corrections. In some states, however, a substantial share of the imprisonment-related expenses are borne by other state agencies or otherwise do not appear in the corrections department’s budget. In order to provide a more complete accounting of the costs of imprisonment, researchers from the Vera Institute of Justice recently collected and analyzed data from forty states (including Wisconsin). Their findings were published in the Federal Sentencing Reporter at 25 Fed. Sent. Rep. 68 (2012).
The Vera researchers identified eleven categories of costs that are not included in corrections budgets. The most important of these, amounting to almost $2 billion in costs nationally in 2010, took the form of gaps in the funding of health benefits for retired corrections employees. In some states, this and other off-the-budget costs added up to a large share of total prison costs. For instance, in both Connecticut and Illinois, about one-third of the total prison cost was outside the corrections budget. When hidden expenses are so high, the public may have a hard time evaluating the true cost-effectiveness of state sentencing and corrections policies.
Wisconsin’s hidden costs, at 8.5 percent of the total, were somewhat below the average among the forty states studied.
Among other things, the Vera report facilitates more reliable cross-state comparisons of spending on imprisonment. For instance, to judge only by corrections budgets, Wisconsin’s spending on prisons seems surprisingly close to Illinois’s, despite Illinois’s much larger size. Once total costs are calculated, however, we can see that our neighbor to the south is actually spending almost exactly twice what we are.
Of course, cross-state comparisons must take into account the relative size of each state’s inmate population. In calculating average cost per inmate, the Vera researchers found surprisingly wide gaps from state to state. At the bottom of the pack were Kentucky ($14,603 per inmate) and Indiana ($14,823). At the top were New York ($60,076) and New Jersey ($54,865). Wisconsin came in a little above average at $37,994 (almost exactly matching Illinois’s $38,268).
As the Vera researchers note, one should not jump to the conclusion that the relatively high-cost states must be badly mismanaged. High costs could be a sign of inefficiency, but, in some cases, they could instead indicate that a state is handling sentencing and/or corrections prudently with an eye to minimizing costs and public-safety threats over the long term. For instance, if low-cost states like Kentucky and Indiana are skimping on mental-health treatment and rehabilitative programming for inmates, they may a high price in recidivism over the long run.
Likewise, low per-inmate costs may indicate that a large share of a state’s inmates are non-violent and require relatively minimal supervision in prison. Such inmates, however, could probably be safely supervised at an even lower cost in the community. Conversely, per-inmate averages effectively penalize the states that do the best job of reserving prison for truly dangerous offenders.
The latter effect may partially explain why Indiana’s per-inmate cost is so much lower than Wisconsin’s. In my ongoing “three states” research (I expect to post an article draft soon), I’ve found notable differences in the composition of the prison populations of Indiana and Wisconsin. While about two-thirds of Wisconsin inmates are being held for violent or sexual offenses, the corresponding number in Indiana is only about forty percent. The Hoosier state, it seems, has a much greater proclivity for imprisoning drug offenders in particular.
In any event, the essential point is this: aggregate and per-inmate imprisonment costs (more accurately calculated now, thanks to the Vera researchers) are a helpful starting point for analysis of a state’s criminal-justice policies, but should not be the ending point — a lot more information may be necessary to contextualize the numbers.