Organizers, scholars, and advocates have long questioned how to increase public support for anti-carceral policies and change the dominant attitudes that uphold support for structural racism. This study examined a deep canvass community organizing tactic that aimed to increase support for a jail decarceration policy proposal in Los Angeles County.

Through a randomized, placebo-controlled field experiment, a community organization had over 1,400 deep canvass conversations with registered voters in majority-white neighborhoods who had previously completed a public opinion survey panel. Two deep canvass conversation models were compared: one that explicitly discussed anti-Black racism in the criminal legal system and another that took a race-absent approach. Outcomes were assessed through participant completion of additional surveys one week and two months after the intervention.

This project addressed several questions related to the intervention’s causal effects: To what extent does deep canvassing increase support for anti-carceral policies and change dominant carceral and racialized attitudes? How do the effects of a race-absent approach compare to those of a race-explicit approach? Are effects durable up to two months after the intervention and after exposure to a counter-message?

Deep canvassing was effective in achieving the community group’s core organizing aims: increasing support for jail decarceration policies and changing dominant carceral attitudes. Both the race-absent and race-explicit approaches were comparable in their effectiveness, meaning that discussing racism did not decrease support for equity policies. Both conditions had limited effects that were durable two months after the intervention and after exposure to a counter-message. Some distinctions did emerge between the race-absent and race-explicit conditions at the level of changing dominant carceral attitudes or issue-specific attitudes––suggesting that there may be some specific ways in which discussing racism or not activates unique considerations that differentially influence attitudes. These distinctions became particularly pronounced over time and after exposure to a counter-message. Considerations for future research and implications for community organizing, abolition praxis, and social welfare are also discussed.