Researchers and industry practitioners have had animated debates on the topic of p-values and statistical significance for decades. In this position paper, Guido W. Imbens outlines the three primary arguments against p-values: 1) p-values do not help decision-makers in contexts where the decision-maker must understand the degree of uncertainty associated with point estimates; 2) p-values may not enable effective comparisons of null and alternative hypotheses; 3) reporting p-values can lead to p-hacking wherein researchers manipulate data and experiment design to achieve statistical significance. Imbens argues that while p-values may be necessary to reject the null hypothesis, they may not help decision-makers evaluating new policies. Instead, he advocates for the application of Bayesian posterior intervals and confidence intervals. Moreover, he recommends pre-analysis plans and other strategies to minimize p-hacking.