The German Socio-Economic Panel (SOEP) is one of the world’s largest and longest-running household surveys, and its personality data have now been analyzed in dozens of papers. Recently, a 2024 article in the Journal of Personality and Social Psychology used SOEP panel data to claim that adult personality continues to change meaningfully with age. The article, Using Within-Person Change in Three Large Panel Studies to Estimate Personality Age Trajectories by Ingo S. Seifert, Julia M. Rohrer, and Stefan C. Schmukle (2024), presents smooth developmental curves and interprets within-person change across three waves as evidence that adults become less neurotic and more mature as they grow older.
This conclusion sounds reassuring. It fits a popular narrative in psychology that personality develops continuously throughout life. It also fits the discipline’s long-standing trust in longitudinal designs as the supposed gold standard for developmental inference. But the conclusion is false. And surprisingly, the evidence for that comes from an earlier, more rigours, analysis of the same SOEP data (Schimmack, 2021).
The real problem is not the data. It is the assumption that changes in panel data can be automatically interpreted as “aging.” This assumption is never stated, never defended, and empirically untenable. When one analyzes the SOEP data with full awareness of the age–period–cohort structure, the exciting story of lifelong personality development evaporates. What remains are large cohort differences, tiny within-person changes, and clear signs that short-interval shifts reflect period effects rather than age.
The key fact is simple: within-person change over time is always a blend of aging and historical period effects. In a short panel study, such as the three-wave SOEP personality assessments in 2005, 2009, and 2013, these two forces move in perfect lockstep. Everyone ages by the same amount between waves, but everyone also moves through identical historical events. And in most societies, year-to-year cultural and psychological shifts are monotonic: economic cycles, mental health literacy, political climate, and cultural norms change in the same direction for everyone, regardless of age. In that situation, the risk of confusing period effects with aging is not just present; it is greater than the risk of confusing age with cohort in cross-sectional studies. Period shifts across eight years are more uniform, stronger, and more directional than cohort differences spread across sixty.
The SOEP data show exactly this pattern. All cohorts move in parallel across the three waves. Young adults, middle-aged adults, and older adults shift by nearly the same amount in the same direction. This is the signature of a period effect. If personality truly changed with age, different cohorts and different ages would show different slopes. They do not. Instead, the 2005–2013 interval appears to reflect historical drift in how Germans respond to personality items, not psychological maturation.
The within-person effect sizes tell the same story. Across the SOEP waves, the typical within-person change is small, often around d = 0.1 or 0.2. These effects are close to noise, especially compared to the much larger cohort differences seen in the same data. For example, the difference in openness between the youngest and oldest cohorts is around d = 0.7. When cross-sectional differences are large but longitudinal changes are tiny, the most plausible explanation is generational or historical change, not adult development. The JPSP article ignores this discrepancy completely.
The deeper issue is one of scientific reasoning. The SOEP reanalysis openly confronts the identification problem: age, period, and cohort cannot be separated statistically without strong external assumptions. The JPSP article makes those assumptions silently. It does not test for period effects. It does not examine cohort-specific slopes. It does not report parallel movement across cohorts. It simply treats the observed change as aging because it comes from a longitudinal design. This is not methodological sophistication; it is a conceptual error disguised by journal prestige.
When scientific rigor and statistical identification are taken seriously, the SOEP data provide no evidence of meaningful age effects in adult personality. The large generational differences observed across decades, and the tiny changes observed within individuals over short intervals, tell us that personality is remarkably stable in adulthood and that apparent “developmental” trends are better explained by history than by age. This evidence is also consistent with results in other longitudinal panels and cross-sectional panels (see Schimmack, 2025, for a review).
The notion of lifelong personality development has intuitive appeal and a long tradition in psychology. But it has always rested on shaky methodological ground. The SOEP data, when analyzed correctly, confirm what many demographers and methodologists have long argued: cross-sectional trends confound age with cohort, and longitudinal trends confound age with period. Without confronting this structure directly, no amount of smoothing, modeling, or journal prestige can turn period effects into genuine aging.
Scientific rigor wins here, not the allure of a developmental story. And the rigorous conclusion is clear: the SOEP data do not show aging effects in adult personality. They show cultural change across generations and historical drift over time. The rest is wishful thinking built on untested assumptions.