The Hungry Judges Study: Why Decision Quality Drops Before Lunch

In 2011, three researchers analyzed 1,112 parole rulings made by eight Israeli judges across ten months. They found one of the most-cited results in modern decision science: the probability of a favorable parole decision started at about 65% just after a food break, fell to nearly 0% just before the next break, then jumped back up after the next break.

The paper became known as the "hungry judges" study. The deeper finding — that decision quality is not constant across the workday and that meal breaks restore it — has been used to justify lunch protection in every serious time-management book since.

Here is the actual study, the methodological debate it generated, and what it means for your afternoon decisions.

The Study

Citation: Shai Danziger (University of Haifa), Jonathan Levav (Stanford), Liora Avnaim-Pesso (Ben-Gurion University), "Extraneous factors in judicial decisions," Proceedings of the National Academy of Sciences, 2011.

URL: https://www.pnas.org/doi/10.1073/pnas.1018033108

The researchers had access to a complete dataset of parole-board rulings in Israel: 1,112 cases, 8 judges, 50 days of court sessions. Each session was divided into three sittings, with food breaks between sittings.

For each ruling, they recorded: time of day, position in the session, prisoner attributes, severity of the original crime, time served, and the binary outcome (parole granted or denied).

The analysis plotted favorable-decision probability against position-within-session. The graph that resulted is the famous one. Probability starts at ~65% at the beginning of each sitting. It drops steadily across the cases heard in that sitting. By the last case before the food break, it is near zero. After the break, it jumps back up.

The pattern repeats across all three daily sittings.

The Interpretation

Danziger and colleagues interpreted the pattern as evidence of "mental depletion" or "decision fatigue." Their reasoning:

• Judges were making long sequences of cognitively demanding decisions.
• The default decision (denying parole) is less mentally demanding than the affirmative decision (granting parole, which requires affirmative justification).
• As fatigue accumulated, judges drifted toward the easier default.
• Food breaks restored cognitive resources, returning judges to the higher baseline.

This is the version of the finding that became famous. It launched a thousand productivity blog posts about decision fatigue and meal breaks.

The Critique

Three years after publication, Andreas Glöckner published a reanalysis in Judgment and Decision Making that challenged the depletion interpretation.

Glöckner's critique:

1. The case-order assumption may not be random. Cases may have been ordered (by judges or court clerks) such that simpler cases came first and harder cases came later within each sitting.
2. If favorable parole was more likely in earlier cases for systemic reasons (not fatigue), the pattern would look the same.
3. Without true randomization of case order, the depletion interpretation is one of several consistent with the data.

The original authors responded with additional analysis defending the original interpretation, but the methodological debate is ongoing.

Bottom line for citing this study: the pattern (decision quality varying systematically across a workday with breaks acting as resets) is robust. The specific mechanism (decision fatigue / ego depletion) is debated. When citing the study, acknowledge both.

Why It Still Matters

The follow-up literature has generated supporting findings using cleaner methodologies:

• Trougakos et al., 2014, Academy of Management Journal: in office workers, higher autonomy during lunch predicted lower end-of-day fatigue and better afternoon well-being. The mechanism doesn't have to be ego depletion; the protective effect of midday recovery is consistent.
• Kühnel & Binnewies, 2014, Journal of Organizational and Occupational Psychology: lunch-break recovery experiences (detachment, relaxation, mastery, control) predicted higher afternoon vigor and self-rated performance.
• Sonnentag detachment framework: across decades of recovery research, psychological detachment from work during off-job time consistently predicts better next-period well-being.

The directional claim — decisions made at the end of a long workday with no break are worse than decisions made after a midday recovery period — survives every reasonable methodological scrutiny.

What does not survive is "lunch breaks improve productivity by X%" — that level of specificity is not supported by any study.

Three Implications for Your Workday

1. The hard decisions belong before lunch, not after

If the Danziger pattern reflects anything real, the worst time to make a hard decision is the period immediately before a missed or skipped lunch. The judges at that point of the day defaulted to "no" — the easier option. Your equivalent: the close-of-day decision when you have not eaten since 7 AM is the worst time to evaluate a major contract, a hiring call, or a strategic option.

The actionable version: if you have an hour-long decision to make, schedule it for 10 AM or for 1:30 PM (after lunch, not before).

2. The reset must be real

Trougakos and Kühnel both find that autonomy during lunch matters more than what you do. A 30-minute lunch where you control the time and choose the activity is more recovery than a 60-minute "lunch" spent answering Slack messages.

The reset is psychological, not just nutritional. Eating a sandwich at your desk while continuing to triage emails does not produce the Danziger reset. Leaving the desk does.

3. The block has to be defended structurally

The Danziger finding describes what happens when meal breaks happen. It does not describe what happens when they don't. The follow-up implication: protect the break.

Most people experience the pattern but cannot name it. The afternoon energy crash, the irritability before dinner, the sense that the last meeting of the day is incoherent — these are downstream of the meal-skipping behavior that has become normal.

A protected lunch hour reverses the pattern. The hour does not have to be perfect. It has to be a real break.

The Calendar-Camouflage Connection

Defending the lunch hour requires that other people respect it. The Danziger finding assumes the meal break happens. The calendar layer is what makes it happen.

A calendar block titled "Lunch" is overridden 38% of the time within four weeks, per our internal research. A block titled like a real meeting is overridden 8% of the time. The Danziger reset only happens if the block holds.

This is what CovertLunch is built for: realistic-looking calendar events written into your lunch window each morning, with daily variation. The Danziger break gets defended automatically.

Try CovertLunch free for 7 days →

The Underlying Principle

The Danziger study landed because it documented something most workers already knew but couldn't articulate: by the end of a long block of decisions, you are not making decisions the same way you were at the start. The judges did not become worse judges as the morning wore on. The cognitive resource they were drawing on depleted, and the easier default became more attractive.

You are not different. Your afternoon meetings, your end-of-day Slack messages, your 4:55 PM "approve or reject" decisions are not happening at the same level as your morning equivalents. The reset window in the middle of the day is the only structural defense.

The reset is named lunch.

Related Reading

• Sonnentag detachment framework applied to lunch
• Microsoft EEG research on the meeting hangover
• Executive calendar management playbook

Frequently Asked Questions

Is the hungry judges study reliable?

The pattern (decision quality drops across a workday and resets with breaks) is well-documented. The specific interpretation (decision fatigue / ego depletion) has methodological challenges. Cite the pattern; be honest about the debate.

Does decision fatigue apply outside parole hearings?

Yes, with caveats. The follow-up research in organizational psychology (Trougakos, Kühnel, Sonnentag) generalizes the pattern to office workers using different methodologies. The specific mechanism may vary; the protective effect of midday recovery is consistent.

Should I avoid important meetings before lunch?

The cleanest reading of the research: schedule consequential decisions for either the start of the day or the period 30–90 minutes after a real lunch break. Avoid the hour immediately before lunch (depleted) and the period 60–120 minutes after lunch (post-lunch dip).

What if I can't take a real lunch?

The Trougakos finding is specific: autonomy matters more than duration. A 25-minute lunch you control is better than a 60-minute "lunch" spent on Slack. Even a 15-minute break away from the desk produces some of the reset, per follow-up Sonnentag research.

How does this connect to CovertLunch?

The Danziger reset only happens if the break happens. The break only happens if coworkers don't book over it. CovertLunch defends the calendar layer so the break can actually occur.