‘If astrological compatibility exists, its effects should be observable’: How one study of 20 million people shows star signs have no influence on romantic compatibility – Image for illustrative purposes only (Image credits: Flickr)
Millions consult horoscopes daily for insights into romance, often pairing signs like fiery Aries with grounded Taurus for supposed perfect harmony. Yet a landmark examination of real-world relationships challenges this notion. Researchers delved into census records from England and Wales, scrutinizing birthdays of over 20 million married individuals to test astrology’s claims on partner choice.[1][2]
The Enduring Appeal of Star Sign Matchmaking
Astrologers have long promoted sun signs as predictors of romantic success. People born under the same zodiac sign or harmonious ones, such as trines at 120 degrees apart, supposedly share traits that foster lasting bonds. Books and websites abound with advice on avoiding squares or oppositions, where signs clash at 90 or 180 degrees.
This belief persists despite skepticism. Surveys show nearly 30 percent of Americans hold some faith in astrology, fueling apps and columns that promise cosmic guidance in love.[3] Proponents argue that even subtle influences from birth months should appear in aggregate data on who marries whom.
Enter the Largest Test of Astrology Ever Conducted
In 2007, David Voas, a senior research fellow at the University of Manchester, seized a unique opportunity. He analyzed data from the 2001 census of England and Wales, which captured birth dates for every resident. This yielded records on more than 10 million married couples, totaling over 20 million people – the biggest sample ever used to probe astrological claims.[2]
Voas sought deviations from random pairing. Under pure chance, the frequency of Aries-Aries couples, for instance, would equal the product of Aries husbands and Aries wives divided by total couples. Any astrological pull should create excesses or deficits in specific combinations. With such scale, even a 0.1 percent effect would produce thousands of anomalous pairs, easily detectable.
Initial Surprises and the Hunt for Explanations
Raw numbers hinted at patterns. Couples sharing the same sun sign appeared about 22,000 more often than expected, with another 5,000 excess for adjacent signs like Aries-Taurus. Same-month birthdays exceeded predictions by roughly 23,000 pairs, and exact shared birthdays ran 41 percent higher – 39,800 observed versus 28,300 anticipated.[1][2]
Voas dug deeper, suspecting artifacts. Census forms, often filled by one household member, invited errors: spouses’ birthdays misentered as the respondent’s own explained same-day spikes. January 1st births ballooned, likely placeholders for unknowns. About 0.5 percent of dates underwent imputation – missing ones assigned as the first of sequential months – artificially boosting adjacent-sign matches.
A pivotal check separated sign from month effects. Pairs with birthdays in the same sign but different months (late Aries with early Aries) showed no excess. Yet same-month pairs across sign boundaries did, pointing to calendar coincidences, not stars. Excluding first-of-month dates and adjusting for errors left a random scatter: deviations under 1 percent, consistent with chance in a vast dataset.
Chi-square tests yielded statistical significance due to sample enormity, but substantive effects vanished. Voas noted, “The couples whose birthdays belonged to the same sign but fell in different months were no more numerous than chance would dictate.”[1]
Why This Matters for Astrology’s Claims
No consistent preferences emerged across 144 sign pairings, regardless of astrological theory. Lacking consensus among astrologers on ideal matches, Voas scanned broadly for any signal – none appeared. Age subgroups (under 40, 40-49, etc.) mirrored the overall randomness.
Critics might invoke full birth charts over sun signs alone. Yet Voas countered that aggregate influences should still surface; sun signs represent the most popular metric. The study focused on enduring marriages, indirectly assessing stability through who stayed wedded.
Key Takeaways from the Census Analysis:
- No excess pairings: Same-sign or “compatible” combinations matched chance after error corrections.
- Error sources identified: Response mistakes and data imputation fully accounted for anomalies.
- Detectable power: Sample size could spot even tiny astrological biases.
- Random romance: Zodiac offered no edge in partner selection or retention.
Lessons from a Data-Driven Romance Audit
Science journalist Carlos Orsi revisited this work in his 2026 book, What Science Says About Astrology, underscoring its rigor. “If astrological compatibility exists, its effects should be observable,” he wrote, echoing Voas’s logic. The findings hold firm nearly two decades later, undimmed by astrology’s cultural sway.[1]
Romantic bonds hinge more on shared values, proximity, and chance encounters than celestial alignments. While stars inspire wonder, this census snapshot reminds us that love defies the zodiac’s script. Future studies might explore modern dating data, but for now, the evidence points squarely to earthly factors.