The Lindy Effect: Why Old Things Keep Lasting
The Lindy effect is the idea that for a non-perishable thing such as a book, an idea, or a technology, expected remaining lifespan grows with current age. A title in print for 40 years is forecast to stay in print roughly another 40. The longer it has survived, the longer it is statistically expected to keep surviving.
This runs exactly opposite to how living things work. A 90-year-old human has a short remaining life expectancy precisely because age signals decay. But a 90-year-old idea carries no biological clock. Its age is not evidence of approaching death; it is evidence of repeated survival against everything that could have killed it. That inversion, that age becomes a credential rather than a countdown, is the whole counterintuitive heart of the concept.
The Lindy's Deli Origin Story
The name comes from Lindy's, a now-defunct delicatessen on Broadway in New York City where comedians and show-business veterans gathered nightly to conduct post-mortems on the latest television comedy. Journalist Albert Goldman sat among them and, on June 13, 1964, published an article in The New Republic titled "Lindy's Law" (pages 34 to 35) describing the folk theorem he heard there.
Here is the twist most people miss: Goldman's original Lindy's Law meant the reverse of the modern version. The deli regulars believed a comedian had a fixed, finite stock of material. The more he spent it through frequent television exposure, the sooner he would burn out. As Goldman put it, the life expectancy of a television comedian was inversely proportional to the total amount of his exposure on the medium, so a performer who rationed himself to occasional specials might last far longer than one who took a weekly slot. Exposure shortened the expected career rather than extending it.
How Mandelbrot Flipped It
About twenty years later, mathematician Benoit Mandelbrot reformulated the idea in his 1982 book The Fractal Geometry of Nature, and in doing so he reversed Goldman's conclusion. Mandelbrot argued that for creative output not subject to a fixed stock of material, every additional appearance increases, not decreases, the expected number of future appearances. The deli philosophers, in his telling, had gotten the sign backward: the more often a comedian had performed, the more performances they were likely to have ahead of them.
Nassim Nicholas Taleb then popularized and broadened the concept in his 2012 book Antifragile, where he first explicitly named it the "Lindy effect." Taleb stripped away the original limitation to a performer's career and extended it to anything without a natural hard upper bound. He stated the rule cleanly: a book in print 40 years should be expected to last another 40, and if it survives a further decade, its expected remaining life resets upward. Mandelbrot, notably, came to agree with Taleb's expanded perishable-versus-nonperishable framing.
Perishable vs. Non-Perishable
The single most important boundary condition is that Lindy applies only to non-perishable things, those with no built-in expiration. Taleb's formulation is precise: for the perishable, every additional day shortens remaining life expectancy; for the non-perishable, every additional day may lengthen it. A loaf of bread, a cat, a star, and a human all have an unavoidable ceiling, so they do not get Lindy. An 80-year-old cannot reasonably expect 80 more years.
Ideas, religions, classic texts, mathematical theorems, and durable technologies are the natural candidates. The QWERTY keyboard, in use since the 1870s, survived the leap from mechanical typewriters to touchscreens and is a textbook Lindy artifact. Euclid's geometry, in continuous use for over two thousand years, is plausibly good for two thousand more. The spoon, essentially unchanged since Roman times, is going nowhere. Crucially, Lindy is a survivorship filter, not a guarantee: most technologies do not last, and the ones that vanish (the fax machine, say) simply never accumulated the age that would have earned them a long forecast.
The Pareto Power-Law Math
Mathematically, Lindy lifetimes follow a Pareto distribution, a heavy-tailed power law. The defining property is a decreasing hazard function: for a Pareto distribution the instantaneous failure rate is proportional to c / t, so the older the thing gets, the lower its momentary chance of dying. Human survival has an increasing hazard; a radioactive atom has a constant one; Lindy things have a falling one.
For a Pareto variable with shape parameter c > 1, the mean lifetime is c / (c - 1). Conditioning on survival to age k, the expected remaining life works out to:
E[remaining | survived to k] = k / (c - 1)
That result is the entire effect in one line: expected additional life is directly proportional to age already lived. Two special cases are worth knowing. In the Mandelbrot and Taleb version (Lindy proportion p = 1), the tail exponent equals 2, the conditional mean is exactly 2 * t0, meaning remaining life equals current age, and the variance is infinite. In the median-doubling version studied by Toby Ord (exponent α = 1), the survival function is the elegant hyperbola S(t) = t0 / t: the chance of lasting past 3 * t0 is exactly 1/3, past 4 * t0 is 1/4, and so on. There both mean and variance are infinite, which is why Ord cautions that such forecasts are driven entirely by an unobservably extreme tail.
Contrast this with the exponential distribution, the memoryless case where remaining life is constant regardless of age. The exponential sits exactly on the boundary: every heavy-tailed distribution with a declining hazard rate displays the Lindy effect, while Lindy is its mirror image where the past actively informs the future. If you want to see the inversion play out on your own example, the Lindy effect calculator projects expected remaining lifespan from an item's current age, and it pairs naturally with related power-law tools like the Zipf's law calculator for word-frequency tails and the Dunbar's number calculator for cognitive limits on stable relationships.
One last honest caveat: Lindy is a heuristic, not a law of physics. It is a robust statistical tendency for things genuinely free of a natural lifespan, but it cannot see a sudden absorbing barrier, the disruptive technology or cultural shock that ends even a long-lived thing overnight. Ord has also shown the effect can arise even from constant or increasing hazard rates, provided there is uncertainty about the size of that rate, so the declining-hazard story is sufficient but not strictly necessary. Used with that humility, Lindy remains one of the sharpest tools for reasoning about what will still be here long after this year's novelties are forgotten.
Frequently Asked Questions
It is the idea that for non-perishable things like books, ideas, and technologies, expected remaining lifespan is proportional to current age. The longer something has already lasted, the longer it is statistically expected to keep lasting.
It comes from Lindy's, a delicatessen on Broadway in New York where comedians gathered. Journalist Albert Goldman wrote up their folk theorem in a 1964 New Republic article titled Lindy's Law, though his original version meant the opposite of today's concept.
People are perishable and have a natural biological upper limit on lifespan, so their hazard rate increases with age. Lindy chiefly applies to non-perishable things with no built-in expiration, where the hazard rate decreases as they survive longer.
Lindy lifetimes follow a Pareto power-law distribution with a decreasing hazard function. For shape parameter c, expected remaining life after surviving to age k is k/(c-1), making future life proportional to age. In Taleb and Mandelbrot's version the tail exponent is 2 and remaining life equals current age.