A COURIER NEWSROOM + DATA FOR PROGRESS EXPERIMENT
A local news article localizing and contextualizing a prominent national issue —access to abortion—written for Americans who are not regular news consumers, successfully changed the minds of those who saw the piece.
Local news holds a powerful place in American public discourse, but it’s national outlets who drive narratives and devour attention. At COURIER, we’ve proven (like we did in Virginia and Iowa) the impact of our trusted local journalism on civic engagement and the power of values-driven coverage of key issues.
During the summer of 2022, COURIER and Data for Progress surveyed, using SMS respondents, approximately 5,200 Michiganders who vote infrequently in elections. The survey covered the issue of abortion, and was conducted in the aftermath of the leaked Dobbs v. Jackson decision published in May 2022.
Given Michigan is one of the American states with what are known as trigger laws — laws that would criminalize abortion immediately upon an overturning of Roe—we conducted the initial wave of our survey specifically on the question of re-criminalizing abortion under the 1931 Michigan trigger law. The survey was conducted from June 14 to June 23. The Supreme Court handed down its official decision on June 24, 2022 which matched the wording and spirit of the leaked decision.
Following the official decision, our partners at Data For Progress randomly assigned voters to either a control or treatment group. The treatment voters were exposed to three weeks of boosted news coverage on Facebook seeing an average of four to five ads each week directing them to an article from COURIER’s Michigan outlet The ’Gander titled, “Six in 10 Michiganders Oppose Re-Criminalizing Abortion. Here’s What One of Them Has to Say.” The control group was withheld from treatment. After the three weeks of ads, treatment and control voters were re-surveyed from July 28 to August 8, asking the same set of questions about abortion rights in Michigan.
Treatment and control voters were asked the following question:
When the Supreme Court legalized abortion in Roe v. Wade, the 1973 ruling that protected women’s right to abortion, it put on hold a 1931 Michigan law that criminalized abortion, even in cases of rape or incest.
If Roe v. Wade is overturned, abortion could be criminalized again in Michigan in all cases except when a woman’s life is endangered.
Would you support or oppose re-criminalizing abortion in Michigan if Roe v Wade is overturned?
Using difference-in-difference analysis, we compared the change in opposition to re-criminalizing abortion over time in the control and treatment groups — using movement in the control group to isolate general effects and as a reference for the treatment group. As demonstrated by the graph below, the increase in net opposition to the question rose by five points in the treatment condition, four points above the control group.
Movement in the control group can be attributed to the change in national discourse about abortion rights — this can also be illustrated by the failure of a re-criminalization effort in Kansas. What is promising about this result, however, is that we can measure movement above what we observe in the control; even when accounting for a changing national discourse, we saw meaningful gains in public opposition to efforts to criminalize abortion.
Voters were also asked (1) whether they support or oppose a lawsuit by Michigan’s Democratic governor, Gretchen Whitmer, asking the state Supreme Court to strike down the 1931 abortion law and recognize the right to abortion as a constitutional right, and (2) whether women who have abortions should or should not be prosecuted if abortion was indeed banned in the state.
We found voters in the treatment group grew to oppose prosecuting women who had abortions by 4 percentage points, and increased support for Governor Whitmer’s lawsuit to prevent the 1931 abortion ban from taking effect by the same margin.
These movements, contrasted with the smaller (or lack of) movement among the control condition indicates that while there was some inherent change in opinion due to exposure to advertising, the treatment of digital ads with links to The ’Gander’s story moved opinion toward greater support for pro-choice positions on prominent issues in Michigan.
Interestingly, the results of this research are consistent with other studies demonstrating that boosted news programs can persuade people to change their minds about highly salient topics, and are an exciting compliment to COURIER’s 2021 research project in Virginia that showed boosted COURIER news could increase turnout in a significant way during a statewide election.
Our hypothesis was that consistent exposure to factual news information about Michigan’s 1931 ban would have a clear effect on public support for or opposition to it.
The results were clear: local journalism can have a real impact on voter sentiment, even at a time of heightened polarization.
Our partners at Data For Progress were responsible for executing the survey and assisting in analyzing the data. The ’Gander article used for this experiment was written by COURIER’s deputy political editor, Keya Vakil.