Charlotte Cordes

Abstract: This study provides a rigorous empirical test for motivated memory as a cause for procrastination. In a longitudinal experiment over four weeks, individuals must complete a cumbersome task of unknown length and receive noisy signals about their assigned workload. Varying scope for motivated memory results in more optimistic beliefs among workers who receive negative signals, but not among those who receive positive ones. This suppression of negative news through motivated memory causes individuals to defer more work to the future.

Abstract: Many important decisions rely on recalled information. This paper investigates whether and how reliance on memory amplifies favoritism in personnel decisions. In an online experiment, managers make incentivized hiring decisions regarding workers of varying quality. Two experimental manipulations isolate the causal effect of memory on favoritism. First, I experimentally vary managers’ preferences for workers. Second, I vary whether managers base their choices on workers' actual quality or on their recall of that quality. When decisions rely on memory, the gap in the hiring probability between favored and non-favored workers more than doubles, from 10 to 21 percentage points. Motivated memory distortions - managers recalling favored workers as higher quality - explain about one third of this effect. The remaining effect is consistent with noisy memory leading managers to hire additional preferred workers with lower recalled quiz scores. Reliance on memory also increases favoritism in recommendations, but not in rankings.

Abstract:  Experimental studies show that individuals update beliefs about ego-relevant information optimistically when they expect no resolution of uncertainty but neutrally when their ability is revealed immediately. This paper studies belief updating and the role of motivated memory when feedback is delayed but eventually disclosed. In a longitudinal experiment, participants receive noisy signals about their relative performance in a IQ-related task (Raven matrices) and learn their true rank four weeks later.  Across subjects, belief updating is asymmetric: unfavorable signals are weighted less than favorable signals. Further, we identify motivated memory among participants who view the task as ego-relevant.

Abstract: Many people struggle to complete long-term projects, such as learning a new language or acquiring a new technical skill. Two major behavioral frictions that prevent the completion of long-term projects are procrastination and forgetting. In theory, procrastination and forgetting can be mitigated with deadlines and reminders, respectively. However, for individuals who procrastinate and forget, only deadlines or only reminders might not suffice to induce project completion. Individuals might still forget to do the task on time (with deadlines only) or still procrastinate endlessly (with reminders only). Based on a theoretical framework, our project empirically assesses the effectiveness of combining reminders and deadlines when multiple tasks need to be completed. We will conduct a field experiment with 4,000 farmers who take mobile-based courses on financial literacy, agribusiness, and farming practices. Participants are randomly assigned to one of five arms: (i) Control, (ii) Reminders, (iii) Deadlines, (iv) Reminders \& Deadlines, and (v) Choice, where participants can self-select into their preferred regime. The experiment will generate detailed data to study the effects on course progression, study dynamics, and learning outcomes.