Being told to “be more empathetic” tends to sound like an appeal to character, as if empathy is something a person either has enough of or doesn’t. A specific meta-analysis, paired with a follow-up brain-imaging study, suggests empathy behaves less like a fixed trait and more like a skill, with real conditions for building it well.
The clearest evidence comes from a 2016 meta-analysis by Emily Teding van Berkhout and John Malouff, published in the Journal of Counseling Psychology. The researchers statistically pooled 18 randomized controlled trials testing structured empathy training programs, comparing people randomly assigned to receive the training against control groups who didn’t, then measuring empathy afterward using standardized assessments. Across the pooled trials, empathy training produced a moderate, statistically reliable increase in measured empathy, adjusted effect size around 0.51 after accounting for potential publication bias. Because these were randomized trials rather than simple before-and-after comparisons, the design supports a real causal claim: deliberately structured practice, not just having the right personality, was what moved the needle.
A second, more detailed study helps explain what’s actually happening inside that training, and adds an important nuance. In a 2014 study by Olga Klimecki, Susanne Leiberg, Matthieu Ricard, and Tania Singer, published in Social Cognitive and Affective Neuroscience, the researchers ran a longitudinal brain-imaging study with 53 female participants, training a subset first in empathy alone and then, sequentially, in compassion, while watching videos of other people suffering both before and after each training phase. Empathy training on its own increased both empathic feeling and negative personal distress, with heightened activity in brain regions associated with feeling another person’s pain almost as one’s own. The subsequent compassion training reversed that distress, bringing negative affect back down to baseline while boosting self-reported positive feeling, with brain activity shifting toward regions tied to reward and social connection rather than personal pain.
Together, these two studies suggest that empathy is genuinely trainable, but that the specific kind of training matters more than simply resolving to feel more. Teding van Berkhout and Malouff’s meta-analysis confirms that structured empathy programs reliably increase measured empathy across a real body of randomized trials, not just anecdotally. Klimecki and colleagues’ research adds a crucial refinement: training in raw empathy alone can increase personal distress right alongside the empathic feeling itself, while adding a compassion-focused component, oriented more toward warmth and care than shared suffering, appears to relieve that distress rather than compound it. The practical takeaway isn’t simply “practice empathy.” It’s that how the practice is structured, whether it leans toward absorbing another person’s pain or toward warm concern for them, meaningfully changes what happens internally.
It’s worth being honest about what these two studies don’t establish. Teding van Berkhout and Malouff’s meta-analysis pools 18 trials that likely varied in the specific training methods used, populations studied, and how empathy was measured, so while the overall effect is real and statistically reliable, it describes an average across meaningfully different programs rather than one single validated technique. Klimecki and colleagues’ study involved a fairly small, exclusively female sample of 53 participants, and it specifically sequenced empathy training before compassion training, so it can’t fully rule out that the order itself, rather than the compassion content specifically, contributed to the shift researchers observed.
Within those honest limits, the research supports a more specific, actionable version of “be empathetic” than the phrase usually implies on its own. Empathy responds to genuine, structured practice the way other skills do, but the research also suggests it’s worth distinguishing between training aimed at feeling someone else’s pain directly and training aimed at warm, caring concern for them, since the second kind appears to build the benefits of empathy without the same risk of quietly wearing a person down.
