Russel Sage Foundation

Co-funded with the JPB Foundation

Criminal justice reform policies seek to change the government’s reliance on carceral institutions, such as police, jails, and prisons, and develop more effective processes to prevent and address crime. Few studies have examined recent reform proposals or assessed interventions to influence attitudes about the criminal legal system. Political scientist Martin Gilens will conduct a randomized, placebo-controlled field experiment to examine the effects of deep canvassing (a method for discussing political issues by sharing personal stories) on preferences towards a decarceration policy, Measure J, and associated criminal justice and racialized attitudes among Los Angeles County registered voters. Passed in November 2020, Measure J is a ballot initiative that allocates 10 percent of unrestricted county funds to community resources and alternatives to incarceration. The study has four randomly assigned treatment conditions: 1) a full deep canvass conversation on Measure J that links the issue to racial inequality and includes both perspective-getting and perspective-taking components; 2) an abbreviated deep canvass conversation on Measure J that does not link it to racial inequality and includes both perspective-getting and perspective-taking components; 3) a full deep canvass conversation on Measure J that does not link it to racial inequality; or 4) a placebo “straw ban” conversation. The experiment would test these hypotheses: 1) Deep canvassing will increase support for carceral reform policies and weaken belief in the fairness of the criminal legal system; 2) Explicitly addressing racialized attitudes will have larger effects than a non-explicitly racial treatment; 3) Deep canvassing will not reduce racial resentment, given that it is a broad and powerful attitude, but linking the issue to racial inequality, by addressing the activation of racialized attitudes through dog whistles, will reduce subtle prejudice –– while the non-explicitly racial treatment will not; and 4) Deep canvassing will increase voter action in support of carceral reform policies, and linking the issue to racial inequality will have larger effects than the non-explicitly racial treatment.