Intuitive decision making has a large and often negative impact in economic decisions, but its measurement and quantification remains challenging. Following research from psychology, behavioral economists have often attempted to causally manipulate the balance of intuition and deliberation by relying on experimental manipulations as cognitive load. However, these attempts have resulted in mixed success, with many null results and no clear general pattern. We explain the possible reasons behind these developments and offer avenues for improvement. First, we show that a very simple formal model of decision processes offers a straightforward test to determine whether cognitive load has been successfully induced, hence disentangling failed inductions and true null results. Specifically, cognitive load in economically-relevant tasks must result in shorter response times. Second, we show that the intuitive arguments on the behavioral implications of cognitive load do not hold on closer, formal examination, unless strong assumptions are made that may or may not hold in typical economic experiments. We then report on seven economic experiments (joint N = 628) using different cognitive load manipulations and confirm the implications of the model. While the effect on response times is strong and pervasive, behavioral effects are weak and elusive. Our research serves as a warning on the differences between economic tasks and psychological experiments and the difficulties associated with importing methods uncritically.