Being the person who always ends up covering, fixing, or finishing what other people leave behind can feel like a fixed part of who you are at work. Research on reciprocity styles suggests it is closer to a missing boundary than a personality trait, and the fix is a specific, repeatable check, not becoming a different kind of person.
Three styles, and a counterintuitive finding
Research by Adam Grant identifies three stable styles people default to at work. Takers, who aim to receive more than they give. Matchers, who reciprocate roughly in kind. Givers, who give more than they receive, sometimes without tracking the cost at all. The counterintuitive part of the finding is where givers actually land in organizations. They occupy both the highest and the lowest performing tiers, while takers and matchers cluster in the middle.
That means giving itself is not the problem. Being a giver is, on the evidence, associated with the strongest long term outcomes available, not just the most depleted ones.
What actually separates the two ends
What separates a top performing giver from a bottom performing one is whether the giving has any boundary at all, not whether they give. The research describes this as otherish giving, generous but not unconditional, against giving with no limit attached to it. The instinct to protect the generosity itself, rather than cut it off entirely, is the more accurate read of what needs to change.
A practical check before saying yes
The most useful operational move from this research is a simple question to ask before accepting an uneven request. Would a fair matcher, someone who reciprocates roughly in kind rather than over giving by default, take this same deal. That question gives an external standard to check against, rather than relying on your own sense of when to say no, which research on this exact population finds tends to run weak.
If a matcher would not accept the terms of a given request, that is useful information, independent of whether you personally feel able to say no in the moment.
Where to start
Before your next request that would mean covering, fixing, or finishing someone else's work, run the matcher test silently before responding. Would someone who reciprocates fairly, not generously, not stingily, take this deal as it stands.
If the answer is no, use that as the basis for a different response, even a partial one, rather than agreeing by default the way you likely have before.
Knowing which specific requests in your own work are actually uneven, versus which ones are reasonable asks that just feel heavy in the moment, is something you will judge more accurately than a general test can on its own. What this gives you is a standard to check that judgment against.
There is a specific way to decline an uneven request that keeps the relationship intact instead of souring it, and it is worth having ready before the next one lands.