Prevailing models typically treat cognitive aging as gradual, with rates of change evolving slowly over time and individual differences expressed mainly as variation in level, onset, and rate of decline. We test this view against an alternative in which individual trajectories are organised as extended plateaus of stability punctuated by brief episodes of accelerated loss. Harmonising 728,000 memory tests from 80,000 participants across three population cohorts, we introduce a simulation-calibrated framework to identify genuine memory stability. 10% of adults ≥ 70 showed stable performance over a decade, and in an MRI subgroup (n ≈ 2,000), these individuals exhibited lower rates of brain atrophy, linking cognitive stability to brain maintenance. However, analyses projecting trajectories backward in time showed that many late-life stable trajectories would imply implausibly low mid-life performance, indicating prior unobserved decline. Across individuals, 54% showed at least one stability period lasting 10 years, but only 0.4% maintained stability over 24 years under our criterion. Models based solely on gradual decline are insufficient to account for the observed temporal structure of memory change, and smooth population-level decline can emerge from punctuated individual dynamics. Together, the findings align with a complex-systems view where stability is often a transient state rather than a lifelong trait.

Det økende kjønnsgapet i barnløshet er gjenstand for betydelig oppmerksomhet fra politikere, forskere og samfunnsdebattanter. En ofte gjentatt påstand er at menn med lav inntekt i økende grad forblir barnløse fordi kvinner «gjenbruker» menn med høyere sosioøkonomisk status. I denne artikkelen benytter vi norske registerdata til å analysere flerpartnerfertilitet – altså å ha fått barn med flere enn én partner – ved 45-årsalder for ulike inntektsgrupper av menn og kvinner født mellom 1945 og 1978. Vi finner ingen tegn til økende gjenbruk av menn med høy inntekt. Flerpartnerfertilitet som fenomen var økende over tid i etterkrigskullene, blant både kvinner og menn, men har siden falt. For alle fødselskull etter 1956 er flerpartnerfertilitet mer utbredt blant kvinner enn blant menn. Blant menn er flerpartnerfertilitet gjennomgående mest utbredt i bunnen av inntektsfordelingen. Videre er veksten i barnløshet blant menn med lav inntekt betydelig større enn det som med rimelighet kan forklares av at andre menn gjenbrukes. Det fremstår ikke som plausibelt at økende flerpartnerfertilitet blant menn øverst i inntektsfordelingen har vært en drivkraft bak økende barnløshet blant menn nederst i inntektsfordelingen.

Are there individuals who resist episodic memory decline into older age? Analyzing 728,000 memory tests from 80,000 participants with at least 4 assessments, we introduce a simulation-calibrated framework to identify genuine memory stability. Across cohorts and models, ~10% of adults ≥70 years showed stable performance over a decade. In an MRI subgroup (n≈2,000), stable performers exhibited lower rates of brain atrophy across widespread regions, anchoring cognitive stability in structural brain maintenance. However, stability was often transient rather than trait-like: many individuals followed trajectories with extended plateaus of stable performance punctuated by episodes of accelerated decline. Accordingly, 54% showed at least one period of observed stability, averaging 10 years, whereas only 0.4% upheld stable performance over 24 years under the strictest definition. These findings are consistent with a complex-systems model of cognitive aging in which decline often reflects critical transitions rather than continuous erosion. ### Competing Interest Statement The authors have declared no competing interest. European Research Council, 283634, 725025, 313440 The Research Council of Norway The Research Council of Norway, https://ror.org/00epmv149, 324882 UiO:Life Science convergence environment Knut and Alice Wallenberg Foundation, https://ror.org/004hzzk67

Science aspires to be cumulative. Reproducibility efforts strengthen science by testing the reliability of published findings, promoting self-correction, and informing policy-making(1). Computational reproductions, whereby independent researchers reproduce the results of published studies, are an essential diagnostic tool(2-10). Such efforts should have greater visibility(11-16). However, little social science reproduction and robustness has been conducted at scale(10,13,17-23). Here we reproduced original analyses and conducted robustness checks of 110 articles that were published in leading economics and political science journals with mandatory data and code sharing policies(17,18). We found that more than 85% of published claims were computationally reproducible. In robustness checks, our reanalyses showed that 72% of statistically significant estimates remain significant and in the same direction, and the median reproduced effect size is nearly the same as the originally published effect size (that is, 99% of the published effect size). Additionally, 6 independent research teams examined 12 pre-specified hypotheses about determinants of robustness. Research teams with more experience found lower levels of robustness, and robustness did not correlate with author characteristics or data availability.

Abstract Lower cognitive ability measured in childhood or late adolescence has been consistently associated with higher mortality risk across adulthood. However, this evidence largely relies on single assessments, leaving it unclear to what extent mortality risk reflects cognitive differences established early in life versus developmental divergence during adolescence—a period of substantial neurocognitive plasticity. Using two nationally representative Swedish cohorts comprising 9,412 males born in 1948 and 1953, we linked cognitive ability assessed in primary school at age 13 years and military conscription at age 18 years to all-cause and cause-specific mortality recorded in nationwide registers through 2025. We decomposed late-adolescent cognitive ability into childhood cognitive level and adolescent cognitive change and evaluated their independent associations with mortality. Childhood cognitive level (HR = 0.81; 95% CI, 0.78–0.85) and adolescent cognitive change (HR = 0.84; 95% CI, 0.79–0.89) independently predicted lower mortality risk, also after adjustment for parental education. Childhood cognitive level and adolescent cognitive change showed partially distinct cause-specific patterns. Childhood cognitive level was most strongly associated with mortality from intrinsic causes, whereas adolescent cognitive change showed relatively stronger associations with external causes, particularly accidental deaths. Although adolescent cognitive change was associated with psychosocial factors including education and psychiatric diagnosis at conscription, its association with mortality persisted after adjustment for these factors. These findings suggest that cognitive development during adolescence carries independent prognostic information regarding long-term survival beyond cognitive level established by late childhood, highlighting adolescence as a consequential period for lifelong health.

Sociological modernization theory predicts that rising meritocracy reduces the influence of family background on education while increasing the role of individual abilities. Sibling correlations are often used as omnibus indicators of family influence, but they reflect both shared environments and shared genetics, complicating interpretation over time. This study examines changes in the genetic and environmental components of educational attainment using administrative data on more than 13,000 Norwegian twin pairs born between 1915 and 1991. We apply a Bayesian hierarchical liability-threshold model to estimate heritability and shared environmental influences across birth cohorts, accounting for changes in educational systems and population distributions. For women, the heritability of educational attainment increased steadily across the twentieth century, while shared environmental influences declined, consistent with expanding educational opportunities and gender equality. For men, neither heritability nor shared environmental influences changed significantly over time. These findings clarify how genetic and environmental contributions to education vary across historical contexts, refining interpretations of long-term trends in family influence.

We use administrative register data from the Norwegian child welfare service (CWS) to analyse the outcomes of 252 144 CWS referrals received between 2014 and 2020. Using a Bayesian hierarchical model, we document extensive and systematic variation across time, across referral sources and across referral concerns. Defining high severity cases as cases in which the CWS seeks mandatory interventions, mandated and nonmandated reporters have similar substantiation rates, but this shared probability declines by half across our data period. The differences in substantiation rates across reporting sources persist when comparing referrals with identical referral concerns. Drawing on the extant literature, we suggest that occupational groups may have evolved diverging reporting practices as a consequence of vague official guidelines combined with a lack of systematic training in identifying relevant cues of maltreatment. We also highlight a possible tension in the Norwegian system: while the family-oriented Norwegian CWS largely opens cases based on referrals, its aim is to work preventively and voluntarily with families before issues reach the severity covered by existing reporting mandates.

Why education is linked to higher cognitive function in aging is fiercely debated. Leading theories propose that education reduces brain decline in aging, enhances tolerance to brain pathology, or that it does not affect cognitive decline but rather reflects higher early-life cognitive function. To test these theories, we analyzed 407.356 episodic memory scores from 170.795 participants > 50 years, alongside 15.157 brain MRIs from 6.472 participants across 33 Western countries. More education was associated with better memory, larger intracranial volume and slightly larger volume of memory-sensitive brain regions. However, education did not protect against age-related decline or weakened effects of brain decline on cognition. The most parsimonious explanation for the results is that the associations reflect factors present early in life, including propensity of individuals with certain traits to pursue more education. While education has numerous benefits, the notion that it provides protection against cognitive or brain decline is not supported.

Although educational attainment is heritable, its conventional measurement in genetic research as years of education (EduYears) is not designed to reveal potential stage-specific genetic influences across discrete milestones. In two Norwegian cohorts (Norwegian Mother, Father and Child Cohort Study, N = 120,527; Norwegian Twin Registry, N = 8,910), we quantified the genetic contributions to completing high school, bachelor’s, master’s and PhD using genome-wide association studies (GWAS), polygenic indices (PGIs) and twin models. Transition-specific analyses, conditioning on prior success, revealed that observed-scale common-variant heritability (h2SNP) and PGI predictability followed an inverse-U pattern, peaking at the transition into higher education (h2SNP ≈ 0.14; R2Tjur ≈ 0.05) before declining for postgraduate degrees. Genetic correlations (rg) with large-scale GWAS of EduYears (EA4) and intelligence (IQ3) were high for early transitions but declined markedly for later ones (e.g., rg with EA4 from ≈ 0.92 to ≈ 0.38). In cumulative analyses, aggregating liability across prior milestones, the gap between twin- and SNP-based heritability narrowed at higher levels of attainment (h2twin ≈ 0.6→0.3; h2SNP ≈ 0.22→0.19), while the genetic overlap between distant milestones diminished (rg ≈ 0.92→0.71). These patterns, obscured by EduYears metrics, highlight a dynamic genetic architecture across educational milestones, refining polygenic prediction and addressing misconceptions about uniform genetic influences on educational progression. ### Competing Interest Statement The authors have declared no competing interest.

We examine long-run effects of automation risk on turnout. We expect gendered negative effects because men's turnout is more sensitive to job loss and earnings, but negative effects might be offset by populist right-wing mobilization on economic grievances. We rely on population-wide administrative data to avoid well-known biases in survey data. We find both men and women with high automation risk to suffer in the labour market, but automation risk is associated with lower turnout for men only. The negative association with turnout is weaker where the populist right is stronger, consistent with mobilization on economic grievances. Finally, we show experimentally that priming of automation risk produces null findings, suggesting that risks need to have material consequences to affect political behaviour. Our findings imply that technological change has contributed to the emergence of gender gaps in turnout and populist voting as well as the participation drop among the working class.

Globalization and automation are leading to skill-biased structural changes in the labor market, resulting in the polarization of employment opportunities. These shifts are raising concerns about growing earnings inequality and gender disparities, particularly in occupations characterized by routine cognitive and physical tasks. This study utilizes comprehensive individual-level data from Norway to analyze gender differences in the routine intensity of occupations. The findings reveal significant and growing gender disparities. These disparities are most pronounced among individuals with low socioeconomic status. The analysis further identifies increasing gender differences in educational attainment as the primary contributor to the growing gender differences. Our results highlight the role of educational inequality in driving labor market disparities, emphasizing the need for targeted policy interventions to address these gendered dynamics, particularly among lower socioeconomic groups.

Why education is linked to higher cognitive function in aging is fiercely debated. Leading theories propose that education reduces brain decline in aging and enhances tolerance to brain pathology or that it does not affect cognitive decline but, rather, reflects higher early-life cognitive function. To test these theories, we analyzed 407,356 episodic memory scores from 170,795 participants older than 50 years, alongside 15,157 brain magnetic resonance imaging scans from 6,472 participants across 33 Western countries. More education was associated with better memory, larger intracranial volume and slightly larger volume of memory-sensitive brain regions. However, education did not protect against age-related decline or weakened effects of brain decline on cognition. The most parsimonious explanation for the results is that the associations reflect factors present early in life, including propensity of individuals with certain traits to pursue more education. Although education has numerous benefits, the notion that it provides protection against cognitive or brain decline is not supported.

Parents and children tend to have similar socioeconomic status (SES). Sociological theory has often emphasized the role of social mechanisms in intergenerational transmission, including the influence of the broader rearing environment as well as parental investments and aid, but often not allotted an important role to genetics. Accumulating evidence suggests that genetics play an important role in the transmission of SES from parents to children. Yet, estimates differ substantially across data sets, measures and methods. Using two research designs that account for potential genetic confounding, and high-quality data from Norway, we estimate the strength of the intergenerational social transmission of a range of SES indicators. By triangulating data and designs, we obtain estimates that are more robust to idiosyncratic modelling assumptions. Measures of Norwegian parents’ socioeconomic position predict their children’s socioeconomic outcomes, but purely social mechanisms only account for a fifth of the total explained variance in intergenerational transmission.

We consider long-term impacts of establishing school psychology offices in Norway. In the 1960s, these offices assessed school readiness resulting in increased postponement of schooling for children born late in the year. We find a positive effect of establishing offices in the 1967-1970 period on adult income, with stronger effects for younger children in each cohort, males, and children in low-income families. After a legislative change in 1976, the offices could reject applications for delayed school entry. In this period, we find no overall benefits of establishing an office, and negative effects on adult earnings for males and disadvantaged children.

We document a convex relationship between earnings rank and cognitive ability for men in Finland and Norway using administrative data on over 350,000 men in each country: the top earnings percentile score on average 1 standard deviation higher than median earners, while median earners score about 0.5 standard deviation higher than the bottom percentile of earners. Top earners also have substantially less variation in cognitive test scores. While some high-scoring men are observed to have very low earnings, the lowest cognitive scores are almost absent among the top earners. Overall, the joint distribution of earnings rank and ability is very similar in Finland and Norway. We find that the slope of the ability curve across earnings ranks is steepest in the upper tail, as is the slope of the earnings curve across cognitive ability. The steep slope of the ability curve across the top earnings percentiles differs markedly from the flat or declining slope recently reported for Sweden.

Young adults from low socioeconomic backgrounds face an increased risk of early mortality. Here we utilize population-wide data from 17 Norwegian birth cohorts (N = 986,573) to assess whether this risk gradient was explained by early-life educational performance, specifically grade point average at 16 years of age. We show that the gradients in both parental education and income largely disappeared when adjusting for school performance in the models. Specifically, among boys, those with the lowest parental education had an unadjusted hazard ratio (HR) of 2.04 (95

Major initiatives are currently attempting to prevent dementia by targeting modifiable risk factors. Low education is frequently pointed to as a potential key factor, due to its robust relationship with dementia risk. Impact of education is notoriously difficult to assess, however, because of associations with multiple other risk and protective factors, and large population-representative samples are required to tease the relationships apart. Here, we studied 207,814 Norwegian men born between 1950 and 1959 who underwent compulsory cognitive testing during military conscription as young adults, to systematically test associations of education, cognition, and other potentially important factors. While low education was associated with increased risk for dementia diagnosis (Hazard ratio [HR] = 1.37, CI: 1.17-1.60), this association was fully explained by earlier cognitive test scores (HR = 1.08, CI: .91-1.28). In contrast, low cognitive score was associated with double risk of later dementia diagnosis, even when taking education into account (HR = 2.00, CI: 1.65-2.42). This relationship survived controlling for early-life socioeconomic status and was replicated within pairs of brothers. The latter finding suggests that genetic and environmental factors shared within families, such as common genetics, parental education, childhood socioeconomic status, or other shared experiences, cannot account for the association. Rather, independent, non-familial factors are more important. In contrast, within-family factors accounted for the relationship between low education and diagnosis risk. In conclusion, implementing measures to increase cognitive function in childhood and adolescence appears to be a more promising strategy for reducing dementia burden.

Offisielle tall basert på standardiserte nasjonale prøver i Norge viser stabile ferdigheter i lesing, regning og engelsk på 5. og 8. trinn mellom 2014 og 2021. Dette er over­raskende: Norske tall viser endring i internasjonale undersøkelser som PISA og TIMSS, pandemien innebar store endringer i skolehverdagen, og lesepraksis og eksponering for engelsk har endret seg systematisk over tid. Vi benytter detaljerte data på oppgave-elev-nivå for å ettergå de offisielle tallene. Prøvene er rapportert på en felles skala ved hjelp av et design der en tilfeldig under­gruppe elever («ankerelever») får deler av oppgavesettet byttet ut med hemmeligholdte oppgaver som brukes over flere år («ankeroppgaver»). Vi viser at de tilfeldig trukne ankerelevene i snitt gjør det like bra som sine medelever på de års-spesifikke oppgavene, men systematisk annerledes enn ankerelever fra andre år når vi ser på sammenfallende ankeroppgaver. Vi analyserer også tidsutviklingen ved hjelp av IRT-modeller, og disse bekrefter klare endringer over tid. På begge klassetrinn lå gjennomsnittseleven i 2021 rundt et halvt standardavvik over 2014-snittet i engelsk. Det er mindre endringer for lesing og regning, men noe dårligere prestasjoner etter 2016. Våre analyser avdekker også årsaken til at de offisielle tallene ikke fanger opp utviklingen. Programvaren Utdanningsdirektoratet har brukt (XCalibre) legger til grunn at ferdighetene hvert år avspeiler en standardisert ferdighetsfordeling. Det er åpenbart umulig å avdekke endring i ferdigheter på tvers av årskull dersom man antar at årskull­ene, i forventning, er helt like. Samlet viser funnene at nasjonale prøver gir en rikere innsikt i elevers ferdighets­utvikling over tid enn tidligere antatt: Nasjonale prøver belyser viktige mønstre av betydning for utformingen av – og prioriteringene i – norsk skole.