The doomscroll tax — how much time you're actually losing

The number people usually quote

If you've read any article on screen time in the last few years, you've seen the headline: adults spend 2.5 hours a day on social media. The number comes from market research surveys, mostly self-reported, and like all self-reported screen time it's a significant undercount. The real measured number from passive-monitoring studies is closer to 3.5 hours a day for the median adult, and 5+ hours for the top quartile.

Most people will dispute this until they actually check. Open Settings → Screen Time on iOS or Digital Wellbeing on Android and look at the last 7 days. The number you see is rarely the number you'd have guessed.

What that buys you

Three and a half hours a day across the median user, spread across Reddit, YouTube, TikTok, X, and Instagram. Call it 1,250 hours a year. That's:

• More time than the median American spends on cooking and eating combined
• About 52 full days of awake time per year
• Roughly 7 years of waking life between ages 18 and 65

These numbers are not moralising. Some of that time is genuinely good — talking to friends, watching things you wanted to watch, learning things you wanted to learn. The question is what fraction of it was that, and what fraction was the algorithm running out the clock on you.

The tax: the share that wasn't worth it

The honest split, based on retention-stream studies in the 2024–2026 attention research:

• ~25% is content you actively wanted (subscribed channels, friends' posts, things you searched for)
• ~35% is content the algorithm chose for you that you would, retrospectively, rate as fine — entertaining, killed some time, no regrets
• ~40% is content the algorithm chose that you would, retrospectively, rate as not worth it — clickbait you fell for, ragebait that made you angry, "engaged" content that made you feel worse rather than better

That last 40% is the tax. At 3.5 hours/day average, it's ~85 minutes/day of attention you'd rather have back. ~14 days a year of awake time.

This is the number a feed filter is actually attacking. Not the screen-time number — the negative-utility share of the screen-time number.

Why "just five minutes" turns into fifty

The mechanism is well-studied. Three things stack:

1. The first scroll is volitional. You opened the app for a specific reason. You found the thing. Fine. 2. The second scroll is habitual. You looked at one more post because the action is muscle memory and costs nothing. 3. The third onward is autoplay. The feed has now had enough signal to know what keeps you scrolling. The algorithm is running. You're not really making decisions, you're sampling outputs of a system that's optimising against your time.

The five-minute estimate is for step 1. Steps 2 and 3 are what produce the 45 unaccounted-for minutes. The fix can't be at step 1 (that's not the problem); it has to be at step 3.

What the maths looks like with a filter

A category-based feed filter doesn't reduce time spent directly. It reduces the proportion of negative-utility content you see. So the time you spend is more often spent on the 25% + 35% you actually wanted, and less often on the 40% you didn't.

Concretely, if the filter cuts the negative-utility share from 40% to 20%, you don't necessarily scroll less — you scroll the same amount, but you walk away rating more of it positively. People who instrument this on themselves usually report two effects:

1. Net screen time drops, by 15–30% — because the algorithm has less to work with, so the rabbit-hole sessions shorten 2. Satisfaction-per-minute rises — they don't feel "I just lost an hour" after a Reddit session as often

The first effect is bonus; the second effect is the actual point. The question wasn't "how do I spend less time on Reddit," it was "how do I not regret the time I do spend on Reddit." Those are different problems with different fixes.

The fix is not "willpower"

The reason willpower-based interventions (Screen Time daily limits, "I'll just check it once") underperform is that they're fighting a system that's evolved specifically to beat them. Every productivity-app and every "30-day digital detox" plays out the same way: works for a week, novelty wears off, you're back where you started. The tools that compound are the ones that change the content of what you see, not the amount of time you let yourself look.

PureFeed is one version of this. There are others. Whichever you pick, the gain is real and measurable — set a baseline by looking at your Screen Time numbers today, install the filter, look again in 30 days. If the numbers don't move, the filter isn't pulling its weight.

The honest endpoint

You're probably not going to stop using social media. Most people who try, fail. The number worth aiming at isn't zero hours on Reddit — it's most of the hours on Reddit being worth it. The 40% tax is the addressable share. The other 60% — the part where you talked to friends, learned something, laughed at a video — was never the problem.

A feed filter is a tax cut, not a moral intervention. It buys you back the part of your scrolling that wasn't worth keeping.