Recency Bias
🇳🇴NærhetseffektDefinition
Recency bias is the tendency to give disproportionate weight to the most recent information, events, or experiences when making judgments and predictions – at the expense of earlier, potentially more representative data. Our memories are not democratic: recent events are vivid, emotionally accessible, and easy to recall, while older information fades into an undifferentiated background. The result is that our perception of reality is systematically skewed toward whatever happened last, regardless of whether it is more informative than what came before.
Real-world example
In performance reviews, recency bias is one of the most documented problems. A study by the Society for Human Resource Management found that managers' annual evaluations correlate far more strongly with the employee's last 2–3 months of performance than with the full 12-month period. An employee who worked excellently for 10 months but had a rough final quarter may receive a mediocre review, while a colleague who coasted for 10 months but excelled recently may be rated as a top performer.
In investing, recency bias drives 'performance chasing': investors pour money into asset classes that have performed well recently (buying high) and flee from those that have performed poorly (selling low) – the exact opposite of rational strategy. After a market crash, recency bias makes investors excessively fearful; after a bull run, it makes them recklessly optimistic.
In sports, a basketball player who misses three consecutive shots is often benched despite having an excellent season average – the coach's decision is driven by recency, not by the much larger sample of the player's overall performance.
In relationships, a single recent argument can overshadow months of harmony, leading one partner to catastrophize: 'We always fight,' when the actual base rate of conflict is low.
Supplementary perspective
Recency bias is one half of the serial position effect (paired with the primacy effect, which gives disproportionate weight to first impressions). It interacts closely with the availability heuristic – recent events are easier to recall, so they feel more representative. It also connects to the peak-end rule, which shows that our memory of experiences is dominated by the most intense moment and the ending. In statistical terms, recency bias is equivalent to using an extremely short moving average to estimate a trend – it captures noise rather than signal. Behavioral economists have shown that even experienced professionals (doctors, financial analysts, sports scouts) fall prey to recency bias unless they use structured decision systems.
Practical advice
Recognize
- —When making a judgment, ask: 'Am I giving this weight because it's recent, or because it's genuinely more informative than older data?'
- —Notice when a single recent event reshapes your entire opinion of a person, trend, or situation.
- —Be suspicious of strong emotions attached to recent experiences – emotional intensity enhances the recency effect.
Counteract
- —Keep written records over time (performance logs, investment journals, decision diaries) and review them before making evaluations.
- —Use structured frameworks with predefined criteria that force consideration of the full time period, not just the latest data points.
- —When evaluating trends, look at data across multiple time horizons (1 month, 6 months, 1 year, 5 years) to check whether the recent trend is representative.
- —In team settings, have multiple evaluators who observed different time periods contribute to assessments.
Ethical use
- —When presenting information to decision-makers, include historical context alongside recent data to prevent recency-driven distortion.
- —Design evaluation systems that weight all time periods appropriately – continuous feedback systems outperform annual reviews partly because they mitigate recency bias.
- —In journalism and communication, contextualize recent events within longer trends to give audiences a more accurate picture.