The past fifteen years have seen the emergence and growth of a bipartisan coalition of policymakers, criminal justice professionals, advocates, and scholars who argue that the American criminal justice system must be reformed. This coalition contends that the nation's practice of "mass incarceration" is costly, inefficient, unsustainable, and the cause of a host of deleterious collateral consequences that disproportionately harm already disadvantaged neighborhoods and the residents therein. Despite justice reinvestment being proven popular among policymakers, little evidence currently exists that measures public opinion about the justice reinvestment paradigm.
Numerous past studies have found that the American public's opinions regarding the criminal justice system varies widely depending on factors related to type of crime, rate of recidivism, and racial identity of the offender. Numerous studies found that exposure to racialized cues (i.e., words, phrases, or images that are closely associated with people of color in American public and political discourse) activates whites' racial stereotypes and causes them to express greater support for punitive criminal punishments and greater opposition to social policies designed to alleviate the material deprivation disproportionately experienced by people of color.
The study had two primary goals:
- To directly assess public opinion about the Justice Reinvestment paradigm of reform, and to measure people's crime prevention preferences within the constraint of a zero-sum budget, as well as their willingness to use alternatives to incarceration for offenders convicted of violent crimes.
- To test whether exposure to a variety of racialized and race-neutral cues significantly affects people's willingness to allocate money into criminal justice institutions versus community-based social services in order to reduce and prevent crime.
A survey experiment was conducted and administered by the firm GfK to participants in its online "Knowledge Panel," a randomly-selected sample of people that is representative of the U.S. population. Survey participants were asked questions similar to those asked of the focus group participants. They were asked to state their support or opposition to sentencing nonviolent offenders to community-based supervision instead of prison, and they were asked to allocate money among six different options in a hypothetical crime prevention budget. The budget options were: hire more police, hire more probation officers, fund clinics to provide healthcare and mental healthcare in communities, increase funding in community public schools, fund economic development programs to create jobs in communities, or give taxpayers a tax rebate.
The rhetoric manipulation of the experiment varied the language used to describe the communities that would receive funding. Each participant was randomly-assigned to one of seven conditions:
- High crime communities,
- High poverty communities,
- Communities with many residents on welfare,
- Inner-city communities,
- Rural communities,
- African American communities, or
- Communities (i.e., the control group).
Prior studies found that the language of conditions 1 through 4 evoke subconscious stereotypes about people of color, making them color-blind but racialized cues. Condition 6 is an explicitly racialized cue, which is hypothesized to evoke less backlash because of social norms against appearing racist. Finally, condition 5 is designated to be a white racialized cue to serve as a point of contrast.
The survey experiment participants were drawn from GfK's Knowledge Panel. The Knowledge Panel is a disproportionate stratified sample of the United States population selected using address-based sampling via the delivery sequence file of the United States Postal Service. A full description of GfK's sampling methodology can be found in the GfK Survey Codebook.
Non-Hispanic White and African-American adults age 18 and over residing in the United States.
Demographic Variables Include:
- Interview start/finish time and duration (minutes)
- Education Level
- Political Party Affiliation
- Household Head, Size, Type, Ownership and Income
- Number of Household Members by Age
- Marital Status
- State/Region of Residence
- Support/Opposition on Sentencing Nonviolvent Offenders
- Budget Surplus Allocation Preferences (Spending surplus on criminal justice vs. community-based programs and services.)
- Voter Behaviors and Attitudes
- Stereotypical Beliefs
- Confidence on Criminal Justice System's Race Equality in Treatment
Variable-Level Statistics for the collection include: 68 variables and 2010 cases
Of the 3916 sampled, 2010 respondents completed the survey - depicting a 51% Survey Response Rate.
Several Likert-type scales were used.