Randomized Controlled Trials (RCTs)
An effective approach to address the challenge of causal attribution involves the utilization of randomization. In this context, a randomized controlled trial (RCT) is employed, wherein a group of individuals is identified as potential participants for a program. They are then randomly divided into one or more "treatment group(s)" (participating in the program) and a "control group" (experiencing no intervention). By adopting this method, it is generally assumed that any significant differences that emerge between the treatment and control groups can be attributed to the program itself.
The randomized controlled trial is widely regarded, including by us, as the benchmark for establishing causal attribution.
In the example of the intervention of sending a letter in the mail with information on how and where to vote to increase voter turnout, and RCT could be conducted by randomly assigning registered voters to either the treatment group (receiving the letter) or the control group (not receiving the letter). Both groups would be exposed to all other election news, advertising and campaign tactics in the same way.
After the election, the number of people who voted in each group (provided by the state’s voter file) would be compared to see if there was a difference. If the treatment group had a significantly higher percentage of people who voted compared to the control group, we can conclude that the letter was effective in increasing voter turnout.