How Does Fiction Reading Influence Empathy? An Experimental Investigation on the Role of Emotional Transportation

Can a good story actually make you a kinder person? Not just for the hour you’re turning pages, but after the book is back on the shelf. Bal and Veltkamp take that simple, sticky question and hang it on one idea: emotional transportation. Think of the times you’ve fallen into a novel and the world around you vanished. That tunnel vision, time-melting focus is what they mean. When transportation is high, your attention, emotions, and imagination align behind the story. When it’s low, you’re skimming. They argue that this immersion is the lever that can shift empathy, and crucially, that the shift doesn’t necessarily happen right away. It can sneak up on you in the days after. Why would fiction do this more than nonfiction? Here they lean on Jerome Bruner’s old distinction. Nonfiction runs on what he called the logico-scientific mode: claims, reasons, and facts lined up for testing. Fiction is the narrative mode. It’s built to make you feel and to simulate another person’s mind and situation in a way that, as Keith Oatley and Richard Gerrig have argued, lets you rehearse social life without the consequences. Fiction also dodges your defenses. If a news story tells you what to feel about a real person, you might resist. If a novel tugs you into someone’s shoes, you just walk. The hitch, Bal and Veltkamp say, is that this only works when you’re actually carried away. Otherwise, there’s no empathy bump. It may even drop. They also float a sleeper effect, which is the idea that the impact of fiction takes time to ripen. You finish a story, and a week later you’re still replaying a scene, revising your sense of a character, pulling a thread through your own life. That’s when change happens. So they designed two experiments to capture that delayed arc and to test the gatekeeping role of transportation. The headline hypothesis is straightforward: fiction can raise empathy over time, but only when you’re emotionally transported into it. Both studies used authentic texts, not lab fables, and both set fiction against a nonfiction control. In the first study, Dutch undergraduates were randomly assigned to read a Sherlock Holmes story or a nonfiction article from De Volkskrant. Sixty-six people took part—thirty-six in the fiction group and thirty in the control—and everyone completed Davis’s empathic concern scale before reading, immediately after, and again a week later. In the second study, a fresh sample of ninety-seven students read a chapter from José Saramago’s Blindness or a set of NRC Handelsblad articles matched for length, again roughly five to five-and-a-half thousand words. The idea was to make both conditions feel like serious reading, the kind you could sink into, and then see whether sinking mattered. To measure that sinking, they used an emotional transportation scale developed by Brian Busselle and Helena Bilandzic. Reliability was solid—mid-eighties in both studies—and it was taken right after reading. They also measured things that could muddy the waters. In both experiments, they asked about narrative understanding and attentional focus to ensure any effect wasn’t just “the fiction was easier to follow” or “I paid more attention.” In the second study, they added a set of emotion checks—happiness, contentment, and awe on the positive side; sadness, anxiety, anger, and fear on the negative—so they could isolate transportation from the raw feeling tone of the reading experience. The empathy measure was the same each time, so they could track change, not just snapshots. The analysis approach will sound familiar if you’ve worked with moderation. They built hierarchical regressions: starting with baseline empathy and the control variables, adding the main effects for condition and transportation, then adding the interaction between the two. Transportation scores were standardized first to keep the math well-behaved. The central test is that interaction term. If it lights up, then the slope linking transportation to later empathy is different for fiction than for nonfiction. And because the whole point is time, they ran the same logic twice—once on empathy immediately after reading and again on empathy a week later. Study 1 sets the pattern. Transportation, on average, didn’t differ between the Sherlock Holmes story and the nonfiction article. People could get absorbed in either one. But attentional focus did run a bit higher in the fiction group, an effect that shows up as an F of five point zero five with a p-value under 0.05. They controlled for that. Now the meaty part. Right after reading, the interaction between transportation and condition didn’t predict empathy. Nothing immediate. One week later, the interaction did. The coefficient on that term was about 0.17 with a p-value below 0.05. When Bal and Veltkamp unpacked it, the simple slope for fiction readers was positive and significant—about 0.09—while the slope for nonfiction readers was flat and not significant. In plain language, the more you were transported by the story, the more your empathy ticked up over the week. If you weren’t transported, it didn’t. And among the least transported fiction readers, empathy actually sagged a bit compared with the nonfiction group. They went a step further with what’s called a region of significance analysis. Imagine a slider for transportation running from low to high. They looked for the points on that slider where fiction and nonfiction meaningfully diverged in their impact on empathy one week later. The safe-for-radio version is this: at the low end of transportation, fiction is not your friend; at the high end, it is. Those contours weren’t explained away by differences in narrative understanding, attentional focus, or by the emotional highs or lows people felt immediately after reading. The effect lived in transportation itself, and it took time to show up. Study 2 repeats the script with new material and tighter controls. Again, transportation levels didn’t differ across fiction and nonfiction on average, and again empathy was assessed before reading, right after, and at a one-week delay. This time, they also included age, gender, and prior narrative experience as covariates, and they explicitly controlled for both positive and negative affect at the moment of measurement. Baseline empathy remained a strong anchor—its coefficient in predicting empathy a week later hovered around 0.81—which is exactly what you’d expect if empathic concern is partly a stable trait. That stability makes any observed change more convincing. The critical pattern reappeared. The interaction between transportation and condition was significant in predicting change from before reading to a week later, with a coefficient around 0.18 and an incremental R-squared bump of about one percentage point. Those are modest numbers; the point isn’t that fiction turns you into a different person, it’s that the direction of change depends on how deeply you were pulled in. When they drew the lines, the slope for nonfiction readers leaned negative, roughly negative 0.12, while the slope for fiction readers leaned positive, around 0.07. That symmetry is telling. Under low transportation, fiction readers drifted down in empathy over the week. As transportation climbed, they drifted up. For nonfiction readers, increasing transportation didn’t produce the same gains and sometimes corresponded to slight declines. They also ran the region of significance logic again to mark out where the fiction and nonfiction curves pulled apart. The gist matched Study 1: the difference emerges at the lower end of transportation and widens as you climb, with the plus side accruing to fiction only when you’ve been drawn in. And all of this held after adjusting for narrative understanding, attentional focus, immediate feelings, and the usual demographic factors. That’s a clean replication with sharper boundaries. Pull these two experiments together, and a clear picture forms. Fiction is not a magic empathy pill. It’s a context. What flips the switch is emotional transportation, and the light turns on slowly. You don’t see it immediately after you close the book; you see it a week later, when the story has had time to be digested and stitched into your sense of other people. That’s what Bal and Veltkamp call an absolute sleeper effect—the main action lives in the delay. And because transportation didn’t differ by condition on average, the story is not “fiction beats nonfiction,” full stop. It’s “when fiction carries you, empathy grows; when it doesn’t, it can backfire.” Nonfiction, even when engaging, didn’t show the same upward tilt. Mechanistically, that maps neatly onto the simulation idea floated by Oatley and Gerrig. When you identify with a character, you rehearse concern for them, weigh their motives, and feel their setbacks. Those micro-rehearsals can nudge empathic concern a bit higher, but only if you actually run the simulation. Skimming won’t do it. And because simulations are sticky—you catch yourself thinking about a scene days later—the effect needs that incubation window. It’s the thinking after that does the work. There are caveats, and they matter. Empathy here is self-reported with Davis’s empathic concern scale. That’s a respectable tool, but it’s still you rating yourself. The samples were Dutch undergraduates, who tend to be pretty empathic to begin with, which raises the possibility of ceiling effects. Both fiction texts were excerpts or single chapters rather than full novels, which is good for experimental control but not always how we consume stories in life. Everything was in Dutch translation, which keeps the language uniform but may sand down some rhetorical edges. Finally, while nonfiction was carefully chosen and matched on length, “nonfiction” is a huge tent; investigative reporting, memoir, and long-form essays can all pull you in very differently. The category label doesn’t do all the work here. Still, the practical message is awfully clear. If you want stories to grow empathy, don’t just hand people any novel and hope. Aim for transportation. That means rich, relatable characters, emotional stakes that feel earned, and enough time and attention to actually get lost. Expect the effect to be delayed. If you’re building a curriculum, schedule the conversation a week later. If you’re designing a reading program, pick works that invite identification instead of just admiration. And be aware that a poorly engaging fiction text can do worse than nothing; it can leave people colder than when they started. The flip side is just as interesting for journalism and science writing. Nonfiction can certainly absorb us, and sometimes it has to. But in these studies, dialing up transportation in nonfiction didn’t produce the same empathy growth. That suggests that if your goal is prosocial feeling rather than knowledge transfer, you may want to borrow more from fiction’s playbook—voice, character, and scene—without bending the facts. Where does this go next? One obvious step is to move beyond self-reports. Imagine pairing these designs with behavioral measures—who volunteers to help, who donates, and who notices subtle emotions in faces—so we can see whether the sleeper effect spills into action. Another is dosage. These were single sittings. What happens with a full novel over two weeks, or with a steady diet of short stories across a semester? The mechanism should be the same, but the magnitude might not be. And then there’s the lever itself. Transportation isn’t just a reader trait; it’s a design choice. Future work can manipulate narrative features—first-person voice, plot twists, and point of view—to causally boost or reduce transportation and watch empathy follow. That would take this from an observed gatekeeping effect to a toolkit. For now, the take-home is deceptively simple. Stories can move us. When they truly do, we move a little closer to each other.