Public and private institutions worldwide have gained considerable traction in developing interventions to alter people’s behaviors in predictable ways without limiting the freedom of choice or significantly changing the incentive structure (Thaler & Sunstein, 2008). A nudge is defined as an intervention to facilitate actions by minimizing friction and removing impediments, while a sludge is defined as an intervention that inhibits actions by increasing friction. While the terms nudge and sludge have garnered significant attention in behavioral economics, psychology, and public policy, the underlying cognitive mechanisms behind these interventions and their impact on behavior change remain largely unknown. We develop a novel cognitive framework by classifying these interventions along six cognitive processes: attention, perception, memory, effort, intrinsic motivation, and extrinsic motivation. In addition, we conduct a meta-analysis of field experiments (i.e., randomized controlled trials) (n=179 papers, k=222 observations, N=4,440,011 participants) from 2008 to 2020 to examine the effect size of these interventions targeting each cognitive process. Our findings demonstrate that effort-reducing interventions (e.g., convenience) are more effective than intrinsic motivation interventions (e.g., commitment-making) to change behaviors. Interventions that reduce or increase friction had similar effect sizes, although there were considerably fewer sludge studies (k=44) conducted to date than nudge ones (k=178). This new meta-analytic framework provides cognitive principles for organizing nudge and sludge with corresponding behavioral impacts. The insights gained from this framework help inform the design and development of future interventions based on cognitive insights.