Abstract The literature on American politics widely agrees that early victories in U.S. presidential primaries are pivotal for securing the nomination, a belief that underpins the front-loading behavior of states. However, demonstrating this success-breeds-success effect is challenging because unobserved candidate qualities could independently link early victories to later success. To address this, we used a regression-discontinuity design, focusing on variations near the victory threshold. Our analysis shows that conclusions about early states rely heavily on limited observations around the cutoff. If any inference is to be drawn, it is that winning in Iowa or New Hampshire has no lasting impact on subsequent contests, nor does winning on any election day affect outcomes on the next. These findings question the presence of momentum effects for winners in the primaries.
Kamphorst et al. (Thu,) studied this question.