Consider an alternative method to estimate causal effects, involving regression analysis between the outcome variable and a binary variable.