The Question the Twins Studies Were Built to Answer
The most fundamental claim of natal astrology — that the positions of celestial bodies at the moment of birth meaningfully shape a person’s character and life trajectory — has a specific, testable implication that doesn’t require any subjectivity, interpretation, or practitioner skill to check. If the claim is true, people born at nearly the same moment should have more similar lives than people born at different moments. The similar birth time means a similar (or identical) astrological chart, which should mean similar astrological influence, which should produce some measurable similarity in the outcomes the tradition claims to predict.
This is what the “time twins” research was designed to test — not whether a skilled astrologer can produce a reading that feels accurate, not whether subjects recognize themselves in astrological descriptions, but whether people with nearly identical birth charts actually show greater objective similarity to each other than to people born at random different times. It bypasses the entire question of practitioner skill, Barnum-effect interpretation, and subjective accuracy assessment by going directly to whether the birth chart correlates with any objectively measurable variable at a population level.
The 1958 Cohort Study
The most substantial time-twins study on record comes from the work of Geoffrey Dean and Ivan Kelly, drawing on data from the UK National Child Development Study — a longitudinal study that tracked a large cohort of children born in a single week in March 1958, following them at ages 11, 16, 23, and 33 and measuring over 100 variables covering cognitive ability, educational achievement, personality, career outcomes, health, and life events.
Dean and Kelly identified all pairs within this dataset born within the same hour of each other — effectively giving them near-identical astrological charts — and compared their similarities across the measured variables against similarities between randomly paired individuals from the same dataset. The sample of “time twins” reached into the thousands, with pairs born within minutes of each other also analyzed as a subset.
The result was unambiguous: time twins showed no greater similarity on any of the over 100 variables than pairs of individuals born on different days. The proximity of birth time — and therefore the near-identity of their astrological charts — predicted nothing about similarity in cognitive ability, educational outcomes, personality measures, career trajectories, health, or life events. The charts were the same. The lives were not.
Why This Is a Stronger Test Than It Might Appear
The time-twins study design avoids several methodological limitations that affect other tests of astrological claims, and it’s worth being specific about why.
Studies relying on self-reported accuracy — asking people whether a reading describes them — are confounded by the Barnum effect, hindsight bias, and all the other mechanisms discussed throughout this series. The time-twins data uses objectively measured outcomes: cognitive tests administered by trained assessors, educational records, health measurements, employment records. There’s no room for the subject to decide whether the “reading” fits.
Studies comparing sun signs to personality measures (discussed in detail in the companion pieces on the Big Five and sun-sign studies) test only a very coarse astrological variable — one-twelfth of the population in the same category. The time-twins study tests something far more fine-grained: people whose entire birth charts, including all planetary positions, house placements, and aspects, are nearly identical. If any version of astrology’s core claim is correct — that the birth chart, not just the sun sign, shapes personality and life — this is the study that should show it.
The sample size — thousands of pairs — provides more than sufficient statistical power to detect even quite small effects. The follow-up period — measurements at four points between ages 11 and 33 — covers the portion of life when astrology would expect chart influence to be most visible. And the range of outcome variables — over 100, covering personality, health, achievement, and life events across multiple domains — ensures that if any astrological effect exists in any of these areas, the study had the scope to find it.
What the Best Astrological Counterarguments Say
The two most serious counterarguments from the astrological community to the time-twins findings are worth examining rather than dismissing, because understanding where they fail is instructive.
The first is that even a few minutes of birth time difference can produce meaningfully different astrological charts — changing rising signs, shifting house cusps, and altering which planetary positions fall in which houses. This argument has some theoretical merit: the ascendant (rising sign) changes every two hours on average, so pairs born within the same hour but with different minutes could have different rising signs in some cases. However, Dean and Kelly’s analysis specifically examined sub-groups of pairs born within narrower windows — within 45 minutes, within 30 minutes, within 15 minutes — and found no increase in similarity even as the birth-time proximity increased and chart similarity became greater. If the argument were correct, the within-15-minute pairs should show more similarity than the within-60-minute pairs. They don’t.
The second counterargument is the “different souls” or “different contexts” argument: that people with the same chart express its potential differently depending on other factors the chart doesn’t capture — soul nature, family context, choices, free will. This is a philosophically coherent position, and it’s one that can’t be directly disproved by the time-twins data. But it comes at a cost: it effectively concedes that the birth chart doesn’t determine or strongly predict life outcomes, which is exactly what astrology’s popular presentation most commonly asserts. A chart that describes “potential” so loosely that people with identical charts can have entirely different lives makes predictions too vague to test — which moves it out of the testable-claim category that Popper’s and Lakatos’s demarcation criteria (discussed in companion pieces) identify as scientific.
What This Doesn’t Settle
The time-twins research is a clean, powerful, methodologically sound finding, and its null result is among the most directly damaging pieces of evidence to natal astrology’s core claim in the empirical literature. It’s worth being honest about what it doesn’t settle, to avoid the “Studies Debunk Astrology Forever” headline problem discussed elsewhere in this series.
The time-twins finding is a strong test of whether birth chart similarity predicts life-outcome similarity at a population level across broad, objectively-measured variables. It’s not a test of whether a skilled practitioner, engaging with a specific individual’s chart in context, can produce insights that genuinely serve that person through some mechanism that wouldn’t show up in population-level statistical comparison. It’s not a test of divination as reflective practice, where the value isn’t prediction but prompted self-examination. And it’s not designed to address the Gauquelin-style planetary-position findings (discussed in the companion meta-analysis piece), which don’t rest on the same “similar charts produce similar lives” structure that the time-twins design directly tests.
What the time-twins finding establishes, cleanly and at scale: people with identical birth charts don’t have more similar lives. If that’s the mechanism astrology proposes, the mechanism doesn’t appear to operate at any detectable strength in real populations followed over decades, across the full range of outcomes the tradition claims to influence.
The Question That Follows
If birth chart similarity doesn’t predict life-outcome similarity, what’s happening when a skilled astrologer reads a specific person’s chart and produces something that feels — to that person, in that moment — genuinely insightful?
The cognitive-bias explanations throughout this series provide a partial account: Barnum statements, cold reading, hindsight bias, and the rest. The neuroscience and psychology articles add another layer: narrative scaffolding, default-mode activation, cognitive-load reduction, self-fulfilling identity effects. What these accounts converge on is a picture in which the reading can be genuinely useful — for reflection, for narrative sense-making, for prompting attention to questions the person hadn’t consciously asked — without the birth chart itself functioning as a predictive instrument.
That picture is compatible with the time-twins null result. It’s not a comfortable picture for the stronger versions of astrological claim. But it’s closer to what the evidence, taken as a whole, actually supports.