Advice to stop comparing yourself to others is useless because comparison is not a decision. Understanding what it is doing makes the situation more tractable.
The basic theory
Social comparison theory proposes that people evaluate their own abilities and opinions by comparing with others, particularly where objective standards are unavailable.
Which is functional. In the absence of an absolute measure of whether you are doing well, others provide the reference.
The process is largely automatic and occurs without intention, which has been demonstrated experimentally.
Upward and downward
Comparison with those doing better generally lowers mood and can raise motivation, depending on whether the gap is perceived as closeable.
Comparison with those doing worse generally raises mood and can reduce motivation.
Which one occurs depends on the situation and on the individual, and people under threat show different patterns from those who are not.
What changed
Historically, the comparison set was people you actually encountered — neighbours, colleagues, people in your town.
Which meant the reference group was roughly comparable in circumstances and included the full range of how people were actually doing.
Now the reference set is global, self-selected and filtered.
You are comparing against the most successful person in any domain you can observe, and against edited presentations rather than lives.
The asymmetry of information
The specific mechanism that makes it worse.
You know your own internal state in full — the doubts, the difficulties, the mundane parts.
You know others through what they present, which is selected and presented deliberately.
Which means every comparison is between your interior and someone else's exterior, and that comparison is structurally unfavourable regardless of the actual facts.
The research on social media
Studies find associations between social media use and negative mood, with effects concentrated in passive consumption rather than active interaction.
Comparison is the most commonly proposed mediating mechanism, and studies measuring it support that.
The overall effect sizes are contested — reanalyses of large datasets have found much smaller effects than headline claims — and the pattern that passive scrolling is worse than interacting is reasonably consistent.
What actually reduces it
Reducing exposure to the comparison, which is more effective than trying to think differently about it, because the process is automatic.
Unfollowing rather than resolving to feel differently, since removing the input removes the comparison.
Comparing against your own past rather than against others, which some research suggests is both more motivating and less damaging.
And noticing the asymmetry explicitly when it happens, which does not stop the comparison and does affect how much weight it carries.
The domain specificity
Comparison hurts most in domains that matter to your self-concept.
Which means someone who does not care about a domain can observe others excelling in it without any effect, and the same person will be affected by a much smaller gap in a domain they identify with.
Broadening what you value, so that no single domain carries the whole self-assessment, appears to reduce vulnerability.
That is easier to state than to do and it does describe what resilient people seem to have.
Where it becomes clinical
Persistent comparison accompanied by low mood, worthlessness or withdrawal warrants attention beyond self-management.
Body image comparison specifically is associated with eating disorders, which are serious conditions with effective treatments and poor outcomes when untreated.
Anyone in that territory should speak to a doctor rather than working on their relationship with social media.
The self-presentation side
Worth considering, since most people are on both sides of it.
Posting a selected version contributes to the environment that affects everyone else, including people you know.
Research on this is limited, and the structural point stands regardless — the comparison problem is collectively produced.
Comparison at work
Where it has practical consequences beyond mood.
Pay comparison specifically has been studied, with evidence that relative pay affects satisfaction more than absolute pay above a threshold.
Which is why pay transparency has complicated effects — it addresses genuine inequity and increases dissatisfaction among those who discover they are below the median, which is by definition half of people.
The research on transparency policies finds effects in both directions depending on implementation.
The useful reframe
Comparison against a specific person doing a specific thing well can be informative if it produces a question about how they did it.
Comparison against a general impression of everyone doing better produces nothing actionable, and that is the version that dominates online.
Curating deliberately
Feeds are algorithmically selected for engagement, and material producing comparison engages effectively.
Which means the environment is optimised against you rather than neutral, and deliberate curation is working against a system rather than exercising simple choice.
Absolute standards
Where an objective measure exists, comparison against it displaces social comparison.
Which is why domains with clear standards — a personal best, a skill test, a measurable output — tend to generate less corrosive comparison than domains where quality is judged only against others.
Creating a personal standard where none exists is one of the few interventions that addresses the mechanism directly.