Case Study 1: The Rise and Fall of Ego Depletion

The Original Claim

In 1998, Roy Baumeister and colleagues published a paper that would become one of the most cited studies in all of social psychology. The claim was elegant and intuitive: willpower is a limited resource, like a muscle. When you use it for one task, you have less available for the next. Resist a plate of cookies? You'll perform worse on a subsequent puzzle. Make a series of difficult decisions? You'll be more impulsive afterward. The energy gets "depleted."

The concept was called ego depletion, and it quickly became one of the most influential ideas in behavioral science.

The Cultural Impact

Ego depletion wasn't just an academic finding. It became a cultural framework:

  • Self-help books advised readers to conserve their willpower for important decisions. "Don't make important choices when you're depleted."
  • Corporate programs restructured decision-making processes around "decision fatigue." "Move your most important meetings to the morning, when willpower is fresh."
  • Parenting advice warned parents that children's willpower is limited and shouldn't be overtaxed.
  • Diet culture adopted the framework: "You only have so much willpower — that's why you binge at night after a day of restraint."
  • Productivity content recommended reducing "trivial decisions" (like what to wear — hence the "Steve Jobs always wore the same outfit" mythology) to conserve willpower for important ones.

The idea was simple, actionable, and felt intuitively correct. Hundreds of studies were published supporting it. Baumeister co-authored a bestselling book about it (Willpower: Rediscovering the Greatest Human Strength, 2011). It was taught in virtually every introductory psychology course.

The Replication Attempt

In 2016, a group of researchers led by Martin Hagger conducted a Registered Replication Report (RRR) — a pre-registered, multi-lab replication of a standard ego depletion experiment. The study involved:

  • 23 laboratories across multiple countries
  • Over 2,000 participants (compared to the original studies, which typically used 30–60)
  • Pre-registered protocols — the methods, hypotheses, and analyses were locked in before data collection began
  • Standardized materials — every lab used the same task and procedure

The result: no significant ego depletion effect. The meta-analytic effect size across all 23 labs was essentially zero (d = 0.04, not significantly different from 0).

What the Meta-Analyses Had Already Hinted

Even before the 2016 RRR, there were warning signs. Carter and McCullough (2014) conducted a meta-analysis of the ego depletion literature and found strong evidence of publication bias — the published studies showed a pattern consistent with selective reporting. When they corrected for publication bias, the estimated true effect shrank dramatically, possibly to zero.

The pattern was clear: hundreds of published studies showed the effect, but the published literature was a biased sample. The unpublished studies — the ones in the file drawers — likely found nothing.

Baumeister's Response

Baumeister disputed the RRR's findings, arguing that the replication used a task that was not sensitive enough to detect the effect, and that ego depletion is real but may operate through different mechanisms than originally proposed. He published a revised theory suggesting that depletion reflects motivational shifts rather than actual resource depletion — people aren't unable to exert self-control, they're unwilling to because they feel entitled to a break.

This revised theory is more defensible but is also a fundamentally different claim from "willpower is a limited resource that gets used up." The shift from "you can't" to "you don't want to" changes the practical implications entirely.

The Aftermath

As of 2026, the scientific consensus on ego depletion is:

  • The original "resource depletion" model — willpower as a limited fuel that gets consumed — is not supported by the large-scale replication evidence
  • There may be motivational or attentional effects that look like depletion in some contexts, but these are much smaller and more nuanced than the original claim
  • The hundreds of published studies supporting ego depletion were likely inflated by publication bias, p-hacking, and small sample sizes

Verdict: "Willpower is a limited resource that gets depleted with use" ❌ DEBUNKED (original resource model) — The 2016 multi-lab Registered Replication Report found no significant ego depletion effect (d = 0.04). Publication bias appears to explain the large published literature supporting the claim. A revised motivational theory may have some merit but makes a fundamentally different claim. Origin: Baumeister et al. (1998). Replication status: Failed to replicate in a pre-registered multi-lab study with 2,000+ participants (Hagger et al., 2016).

The Lesson

Ego depletion is a perfect case study in how the replication crisis works:

  1. An intuitive, appealing idea (willpower as a muscle) is proposed
  2. Small-sample studies appear to support it
  3. Publication bias filters out null results, creating an artificially strong evidence base
  4. The idea goes viral — books, TED talks, corporate training, self-help advice
  5. A large, pre-registered replication finds essentially nothing
  6. The viral version persists long after the scientific support has evaporated

The ego depletion story illustrates why the replication crisis matters for everyday life. People made real decisions — about their schedules, their diets, their parenting, their work habits — based on a finding that turned out to be unsupported. The self-help industry built an entire framework around "conserving willpower." That framework doesn't have the evidential foundation its proponents claimed.

Discussion Questions

  1. If you previously structured your day to "conserve willpower" (morning decisions, evening relaxation), does the ego depletion replication failure mean you should change your routine? Or could the routine be beneficial for other reasons?

  2. The revised motivational theory suggests people aren't depleted but feel entitled to a break after exerting self-control. Does this distinction matter practically? How would advice differ under each model?

  3. Over 200 studies were published supporting ego depletion before the RRR. How should we evaluate a large body of literature that may have been built on publication bias? Can we trust a finding just because "hundreds of studies support it"?

  4. Baumeister's Willpower book (2011) sold hundreds of thousands of copies. What responsibility do scientists have when their popularized findings fail to replicate? Should the book be updated, recalled, or simply allowed to remain in circulation with a note?