Why Does Every Year Feel Faster Than the Last?

Because the length a year has in hindsight is not read off a clock — it is rebuilt from whatever distinct, retrievable material the year left behind, and a year made mostly of familiar weeks leaves less of that material than a year made of first times. That is the mechanism in one line: remembered duration is a function of memory density, not of elapsed time, which is why a year can be exactly 365 days long and still arrive in November feeling as though it skipped a queue.

Two caveats before the evidence. The research below describes averages in particular samples, with effects that are real and modest — none of it predicts how your own year will feel. And the tidiest version of this story, the one you have probably heard, turns out to be the least well supported part of it.

Two different questions that both sound like "how long was that?"

Time perception research splits the question in a way ordinary language does not.

In a prospective judgement, you know in advance that you will be asked about duration, so you attend to time while it passes and encode it deliberately. In a retrospective judgement, you find out only afterwards, so any temporal information was picked up incidentally, and the estimate has to be reconstructed from what happens to be in memory.

Richard Block and Dan Zakay's meta-analytic review in Psychonomic Bulletin & Review in 1997 pooled 20 experiments and found the two do not agree: prospective judgements were significantly longer than retrospective ones, with a weighted mean effect size of d = 0.35, a duration judgement ratio about 16% greater, and roughly 15% more variability between people in the retrospective condition. Their account of what drives the retrospective estimate is the load-bearing part for this article — remembered duration rises with "the amount of stored and retrieved information," "the number of remembered changes," or "the number of encoded and available contextual changes."

One qualifier stops that from becoming glib: in their analysis stimulus complexity influenced remembered duration while the sheer number of stimuli did not. Density is not a headcount of events. It is how much distinguishable material survives.

A more recent study adds a layer. Natalia Martinelli and Sylvie Droit-Volet, in Scientific Reports in December 2022 with 803 undergraduates, separated a duration judgement (how long was that?) from a passage-of-time judgement (did that feel fast or slow?). At durations of seconds the two were unrelated, and attention-related factors predicted the duration estimate. At durations of minutes, attention stopped being the main predictor, emotion — boredom and sadness — took over, and the two judgements became correlated.

So "the year went fast" and "the year was short" are not the same claim, and they only partly travel together.

What the age research actually shows

The most-cited study on this is Marc Wittmann and Sandra Lehnhoff's, published in Psychological Reports in 2005, using a battery of questionnaires with 499 participants aged 14 to 94. Its conclusion is genuinely supportive of the folk belief — and considerably more hedged than the folk belief:

> "the momentary perception of the passage of time and the retrospective judgment of past periods of time are a function of chronological age; however, small-to-moderate effects accounted for at most 10% of the variance."

At most a tenth of the variance. That is a real signal and a small one. Age is one input among many, not the switch.

Two further details matter more than the headline. The effect is not uniform across intervals: as a 2026 paper summarising this literature puts it, "the effect is specific to the period of the last 3 to 10 years; no significant associations are observed for shorter periods, such as the previous week, month, or year." And it does not run away forever — there is "evidence that the effect of age on the perceived speed of time over the past 10 years may reach a plateau at around the age of 60."

This is where the famous explanation falls down. The "proportional theory" — that a year feels shorter because it is a smaller fraction of the life already lived — was first formulated by Paul Janet in 1877. It is a nineteenth-century speculation, not a finding, and it has never been the conclusion of a study. It also makes the wrong predictions: a shrinking proportion implies smooth, continuing acceleration, and should apply to a month as much as to a decade. The measured pattern is lumpier — interval-specific, and flattening in later life.

The 2026 study that separates two explanations

If not proportion, then what? A study published in April 2026 in Memory & Cognition was built to pull two candidate mechanisms apart. Alice Teghil, Sebastian Wittmann, Adele Lifrieri, Sophia Saad, Maddalena Boccia and Marc Wittmann tested 120 adults (60 men, 60 women) aged 20 to 91, in six age bands, using a Subjective Time Questionnaire, an autobiographical fluency task and a cognitive screening measure.

The results, as reported:

- Older participants rated the past decade as having passed faster (r = .26, p = .004) — but not the past year, reproducing the interval specificity above.

- The memory load account failed. Perceived speed over the decade did not correlate with the quantity of autobiographical memories retrieved (p = .305), nor with how participants rated those memories for significance, emotional valence or vividness (all p > .05).

- What did predict it was encoding: lower delayed memory recall performance predicted faster subjective passage (p = .001), and reduced capacity to encode new information partially mediated the age effect. Overall cognitive screening scores were negatively correlated with perceived acceleration (p = .004).

Read plainly: in this sample, the decade felt faster not for people who had accumulated more life, but for people laying down less new material as they went. The title says it — memory encoding for new information, not autobiographical memory load.

Hold that at the right strength. It is 120 people, measured once, correlational; it cannot settle which way the causal arrow runs, and it is a pattern across a sample rather than a statement about anyone's memory. But it is the cleanest available test of the two explanations, and it points away from "more years to divide by" and towards "how much of this year was new."

Why some months vanish and others do not

The other way to test the idea is to find a period when everybody's clock time was identical and their experience was not. The 2020 lockdowns were exactly that, and Ruth Ogden studied it in PLOS ONE in July 2020 with 604 UK participants. More than 80% reported time felt distorted compared with normal — and the shape of the distortion is the striking part:

How UK participants rated the passage of a day and of a week during the first Covid-19 lockdown, compared with normal. Ogden (2020), PLOS ONE, n = 604.*

The same calendar, the same restrictions, the same number of hours — and the sample split almost down the middle. For the passage of a day, 40% said slower than normal and 41% said faster. For a week, 39% slower and 48% faster. In Ogden's analysis, slower passage was associated with increasing age, increasing stress, reduced task load and reduced satisfaction with social interaction.

A second study complicates the neat version, and it should not be left out. Ogden and Andrea Piovesan returned in PLOS ONE in July 2022, asking 766 people (mean age 28) to judge the preceding 12 months around the one-year anniversary. Only 9% said the year felt like 12 months; 57% said it felt longer than 12 months, and 34% shorter. Overestimation was associated with greater depression and anxiety, reduced satisfaction with social interaction, lower physical activity, greater changes to daily routines, and having been advised to shield.

So a disrupted year did not simply compress. For most of that sample it stretched — and the stretching travelled with distress, not only with novelty. Memory density is a decent account of why a blurred year is hard to navigate in hindsight; it is not a licence to claim that any eventful period feels long or any quiet one short. Emotion sits in the middle of this, exactly as Martinelli and Droit-Volet found at the scale of minutes.

The reminiscence bump, and what it does not prove

The last piece usually bolted onto this argument is the reminiscence bump: across studies, older adults recall a disproportionate number of personal events from roughly ages 11 to 25. The bump is one of the most robust findings in autobiographical memory research. The evidence for the usual explanation — that the period is full of first times, so it is densely encoded — is much weaker.

Tabea Wolf and Daniel Zimprich, in Memory & Cognition in 2020, modelled 2,813 memories reported by 97 adults aged 60 to 88 in response to 31 cue words. First-time experiences did mark out bump memories — 49.8% of them — but they were even more likely to come from the period before the bump. The authors' conclusion: an account focused only on novelty "can explain why memories from the reminiscence bump differ from memories from succeeding life periods. However, it is not suited to explain the increase in the number of memories from the second decade of life." Emotional valence was the stronger predictor.

Steve Janssen and Jaap Murre went further in the Quarterly Journal of Experimental Psychology in 2008, with an internet study of nearly 3,500 participants, reporting that the bump consisted of relatively fewer novel, emotional and important events than neighbouring periods. They proposed instead that memory is simply enhanced in adolescence and early adulthood — an explanation they note has not been directly tested.

The honest summary: novelty is a real ingredient in what separates a densely remembered stretch from a thin one, and it is not a complete theory of anything. Enough for the practical part; not enough for the sweeping part.

The small practical bit

This is deliberately short, because the evidence supports three modest moves and not a philosophy.

1. Put first times in on purpose. Not big ones. A route you have not walked, a dish you have not cooked, a person you have not seen in two years. First-time experiences are what distinguished bump memories from later-life memories in Wolf and Zimprich's model, and they are the one input you can actually schedule.

2. Mark the transitions instead of letting weeks butt up against each other. Ashley Bangert and colleagues, in Attention, Perception, & Psychophysics in 2020, found that tones separated by an event boundary in a film were judged as farther apart than tones inside the same event. That is seconds in a lab, not months in a life — a nearby result rather than proof — but the direction fits. In practice: change the location, the company or the format at deliberate points, so a month has edges.

3. Record something dateable. One line a day with the date on it, or a photograph a week filed by month. If the hindsight estimate is rebuilt from retrievable material, the cheapest intervention is leaving more of it behind. Least glamorous item here, probably the most effective.

What this is not: an argument about mortality, or a prompt to live more fully. The finding is narrow. A year with more distinguishable material in it is easier to reconstruct as long, and that is all it says.

Frequently asked questions

Is it true that time speeds up as you get older?

On average, and modestly. Wittmann and Lehnhoff found the pattern across 499 people aged 14 to 94, but the effects accounted for at most 10% of the variance. It is also interval-specific: the association shows up for judgements of the last 3 to 10 years and not for the previous week, month or year. And it appears to level off — there is evidence of a plateau around age 60 rather than continuous acceleration.

Does a year feel faster because it is a smaller fraction of my life?

That is the proportional idea, first put forward by Paul Janet in 1877. It is a nineteenth-century proposal rather than a research finding, and it is worth knowing that it makes predictions the data do not match: it implies smooth, continuing acceleration and should apply to short intervals as much as long ones, whereas the measured effect is specific to multi-year periods and appears to flatten in later life.

So what does drive it, if not proportion?

The best current test points at encoding. In the April 2026 Memory & Cognition study of 120 adults aged 20 to 91, perceived speed of the past decade was unrelated to how many autobiographical memories people could retrieve, but was predicted by delayed recall performance — the capacity to lay down new information. That is one correlational study, and it describes a sample rather than any individual, but it separates two explanations that usually get bundled together.

Why did lockdown feel both endless and like it disappeared?

Because those are two different judgements. Ogden's 2020 study of 604 UK participants found more than 80% experienced distortion, split almost evenly: for the passage of a day, 40% said slower than normal and 41% faster. A year later, Ogden and Piovesan found 57% of 766 people felt the preceding 12 months had lasted longer than 12 months. How a period feels while it passes and how long it measures in hindsight can move in opposite directions.

Does being busy make a year feel longer or shorter?

The research does not support a clean answer, which is why Mind Decoded is not giving you one. Block and Zakay's meta-analysis found remembered duration rises with the amount of stored and retrieved information and with remembered changes — but also that stimulus complexity mattered while the sheer number of stimuli did not. Busy-but-identical is not the same input as varied. Emotion complicates it further: at longer durations, boredom and sadness predicted the felt passage of time more strongly than attention did.

Should I be worried if this year vanished on me?

No, and this article is not a basis for concluding anything about yourself. Everything above describes averages across samples, measured with modest effects, using group-level questionnaires that were never designed as individual assessments. A year that is hard to reconstruct is most often a year that was structurally similar week to week. If you have a specific concern about your memory, that is a conversation for a doctor, not a blog post.

Connect

Questions or thoughts? Reach out anytime.

Email

subscribe to our newsletter

hello@minddecoded.com

© 2026. All rights reserved.

Terms of Service Privacy Policy Disclaimer