No-Hair Theorem

The statement and what it forbids

The no-hair theorem (more properly the black-hole uniqueness theorem) is the strongest classification result in classical gravity. It says that any stationary, asymptotically-flat, regular, electrovac solution of the Einstein-Maxwell equations whose exterior contains a regular event horizon belongs to a single four-dimensional family — the Kerr-Newman family — parameterised by three real numbers: gravitational mass M, electric charge Q, and angular momentum J. (Magnetic monopole charge P is a possible fourth parameter in extensions; in the standard astrophysical setting P = 0.)

Everything else that a star or other progenitor might have had — composition, internal pressure profile, baryon number, lepton number, magnetic moment beyond what the spinning charge produces, higher multipole moments of the mass distribution, asymmetries of any kind — is absent from the final black-hole solution. Those properties are not preserved as "hidden" parameters inside the horizon; they are radiated away as gravitational and electromagnetic waves during the collapse and the ringdown that follows. The horizon is a kind of perfect amnesiac: it remembers only how heavy, how charged, and how rapidly spinning the things that fell in were, integrated over all infalling matter.

Two stationary black holes with the same (M, Q, J) are physically identical outside their horizons. There is no measurement an external observer can perform that distinguishes them. This is the precise content of "no hair".

Wheeler's slogan and the metaphor

In 1971 John Archibald Wheeler popularised the phrase "black holes have no hair" after Remo Ruffini's earlier remark that the cosmetic features of the collapsing matter would be "shaved off" at the horizon. Wheeler chose the metaphor for vividness: a head of hair is the visible mark of individuality on an otherwise round object, and a Schwarzschild black hole is exactly such a "bald" round object whose parameters reduce to a single number. Adding charge and spin gives at most three. The metaphor was risky enough that it took years before the technical journals adopted it; today it is universal.

The slogan was timed well. In 1963 Roy Kerr had written down the rotating black-hole solution; in 1965 Newman et al. had extended it to a charged spinning case. By 1971 there was a one-parameter (Schwarzschild), two-parameter (Kerr, Reissner-Nordström), and three-parameter (Kerr-Newman) family of explicit solutions, and the question was whether anything else could be a black hole. Wheeler's slogan stated the answer as a guess; the Israel-Carter-Robinson program supplied the proof.

The three uniqueness proofs

The classical theorem was assembled in three steps over eight years:

Year	Author	Result
1967	Werner Israel	Schwarzschild uniqueness: any static, asymptotically flat vacuum black-hole exterior is Schwarzschild (one parameter, M).
1968	Werner Israel	Reissner-Nordström uniqueness: any static, asymptotically flat electrovac black-hole exterior is Reissner-Nordström (two parameters, M and Q).
1971	Brandon Carter	Kerr uniqueness (axisymmetric case): any stationary, axisymmetric, vacuum black-hole exterior is Kerr (two parameters, M and J).
1972	Stephen Hawking	Rigidity theorem: any stationary black hole must be axisymmetric (the timelike Killing vector becomes null at the horizon, generating a second Killing vector).
1975	David Robinson	Closure of Carter's argument: removing residual gaps in the Kerr-uniqueness proof.
1980-90s	Mazur, Bunting, Wald	Extensions to Kerr-Newman (three parameters M, Q, J) and tightening of regularity assumptions.

The chain Israel → Hawking → Carter → Robinson is the classical theorem. The 1980s and 90s saw extensions to higher dimensions, modified gravity sectors, and careful relaxation of the analyticity assumption (which was a weak point of Hawking's rigidity proof until Friedrich, Rácz and Wald addressed it in 1999).

The Kerr-Newman family

The most general stationary, asymptotically-flat, electrovac black hole is described by the Kerr-Newman metric, which in Boyer-Lindquist coordinates can be written

ds² = -(Δ/Σ)(dt - a sin²θ dφ)² + (sin²θ/Σ)((r² + a²) dφ - a dt)² + (Σ/Δ) dr² + Σ dθ²

with Σ = r² + a² cos²θ,   Δ = r² - 2Mr + a² + Q²,   a = J/M

Three parameters appear: M (mass), Q (charge), a = J/M (specific angular momentum). All other features — the location of the horizons, the ergoregion, the singularity structure, the multipole moments — are functions of these three numbers alone.

The horizon exists only when M² ≥ a² + Q² (cosmic censorship); the limit M² = a² + Q² is the extremal Kerr-Newman black hole. Most astrophysical black holes have Q ≈ 0 and a/M ≲ 0.998 (the Thorne limit set by accretion-disk torques), so the two-parameter Kerr surface (M, J) suffices in practice.

The multipole hierarchy

One way to state the no-hair theorem precisely is in terms of multipole moments. Any asymptotically-flat stationary spacetime has a tower of mass moments M_ℓ and current moments S_ℓ (Geroch 1970; Hansen 1974). For a generic source these moments are independent. For a Kerr black hole, however, they obey

M_ℓ + i S_ℓ = M (iJ/M)^ℓ

Every higher multipole is fixed by M and J. In particular the quadrupole moment is Q_22 = -J²/M, the octupole S_3 = -J³/M², and so on. Any non-Kerr compact object — a boson star, a wormhole, a black hole with hair from a scalar coupling — would have free higher moments and would betray itself in an extreme-mass-ratio inspiral or in a sufficiently detailed ringdown spectrum. Confirming the multipole hierarchy is sometimes called "Kerr spectroscopy".

Quasinormal modes and the LIGO test

The empirical handle on no-hair is the ringdown of a perturbed Kerr black hole. A small perturbation — for instance, the post-merger remnant of two coalescing black holes — decays exponentially through a discrete spectrum of damped sinusoids called quasinormal modes (QNMs). Each mode is labelled by integers (ℓ, m, n) and has a complex frequency ω_ℓmn = (real part) + i (damping). For a Kerr hole, all ω_ℓmn depend on (M, J) and on nothing else.

The cleanest no-hair test is therefore: from the inspiral and merger of a binary-BH gravitational-wave signal, infer the remnant (M_f, J_f). Then measure two or more QNM frequencies from the post-merger waveform, and check that their (M, J) inferences agree. Any extra mode parameterised by an unrelated frequency would constitute hair — non-Kerr behaviour.

Event	Year	Remnant mass	Remnant spin	No-hair test
GW150914	2015	62 M☉	a/M ≈ 0.67	Dominant 220 mode detected; consistent with Kerr
GW190521	2019	142 M☉	a/M ≈ 0.71	Marginal evidence for 220 + 330 multi-mode
GW200129	2020	61 M☉	a/M ≈ 0.74	High-SNR ringdown; mode consistency at the few-percent level
GW230529	2023	~6 M☉ NSBH	a/M ≈ 0.2	Lower SNR ringdown; mode-frequency tests in progress

The LIGO-Virgo-KAGRA Testing GR papers (2016, 2019, 2021, 2024) report no statistically significant deviation from the Kerr-spectroscopy prediction at current sensitivity. Future detectors (Einstein Telescope, Cosmic Explorer, LISA) are expected to measure four or more QNMs per event and to push the no-hair test to part-per-thousand precision.

Worked example: Schwarzschild radius and ringdown for 10 M☉

Take a 10 M☉ stellar-mass black hole formed by a binary merger. Its Schwarzschild radius is

r_s = 2GM/c² = 2 × 6.674e-11 × 1.989e31 / (3e8)²
     = 2.95e4 m × (M / M☉) for non-rotating
     ≈ 29.5 km for 10 M☉

For a single solar mass the canonical value is the often-quoted Schwarzschild radius for solar mass: 2.95 km. The light-crossing time of the horizon is t_g = r_s/c ≈ 0.1 ms for 10 M☉. The dominant (ℓ = 2, m = 2, n = 0) quasinormal mode of a Schwarzschild hole has frequency

M ω_220 ≈ 0.3737 - 0.0890 i  (geometric units)
f_220 ≈ 12 kHz × (M☉ / M)
     ≈ 1.2 kHz for 10 M☉
τ_220 ≈ 0.55 ms × (M / M☉)
     ≈ 5.5 ms for 10 M☉

So a 10 M☉ post-merger remnant rings at about 1 kHz with a decay time of a few milliseconds. For LIGO's 10 Hz - 5 kHz band this is squarely in the sensitive region. Higher-spin holes shift these numbers: at a/M = 0.7, ω_220 ≈ 0.532 - 0.083 i, raising the frequency by ~40% and slightly lengthening the decay. Detecting two QNMs and confirming they agree on (M, J) is the no-hair test in operation.

How exactly is the hair lost?

The dynamical mechanism is gravitational radiation. When a deformed, irregular distribution of matter collapses to form a black hole, the resulting spacetime is initially non-stationary — it carries multipole moments inconsistent with the no-hair endpoint. Those non-stationary moments couple to gravitational radiation and are emitted on the dynamical timescale t_g ≈ r_s/c. Higher multipoles are radiated faster than lower ones, so the configuration settles down through a brief, damped ringdown phase in which the hole "sheds its hair" outward as gravitational waves.

For a charged collapse the analogous shedding occurs in the electromagnetic sector: higher electromagnetic multipoles are radiated as light. In both cases the energy budget is modest — at most a few percent of the rest-mass energy of the infalling matter — but the information content carried away is large. The exterior spacetime relaxes to Kerr-Newman in a few light-crossing times of the horizon.

The famous result of Price (1972) made this precise: a generic perturbation of Schwarzschild decays as a power law t^(-(2ℓ+3)) for the multipole moment ℓ. Higher moments fade faster. The end state is bald.

What no-hair does not forbid

• Temporary dressing. A real astrophysical black hole carries an accretion disk, magnetic field lines, jets, and tidally-induced bulges from companions. These are properties of the environment, not the hole. The no-hair theorem only constrains the stationary vacuum/electrovac exterior of an isolated hole.
• Multipoles fixed by (M, J). A Kerr hole has a quadrupole, an octupole, and arbitrarily high moments — they are simply functions of (M, J), not free.
• Quantum hair. Recent work (Hawking, Perry, Strominger 2016; Strominger 2018) argues for "soft hair": large-gauge transformations and supertranslations at infinity that imprint zero-energy charges on the horizon. These are quantum, infrared, and do not violate the classical theorem.
• Modified-gravity hair. Theorems in scalar-tensor, Einstein-Maxwell-dilaton, and higher-curvature gravity allow stationary "hairy" solutions. The classical no-hair theorem is a statement about pure Einstein-Maxwell, not about gravity in general.
• Higher dimensions. In five or more dimensions, black-string and black-ring solutions exist with mass, spin, and possibly other independent topological charges; the four-dimensional uniqueness theorem does not extend.

Variants and extensions

• Magnetic monopole hair. A black hole can in principle carry magnetic charge P, giving a four-parameter family (M, Q, P, J). Dyonic Kerr-Newman is well-studied but astrophysically irrelevant (no monopoles observed).
• Scalar hair in modified gravity. Dilatons, axions, and many other scalars can dress a black hole if the underlying Lagrangian permits. Such hair is a smoking gun for new physics; current LIGO constraints place upper bounds on dimensionless coupling parameters at the 10⁻² level for some models.
• Yang-Mills hair. Bizon (1990) and Volkov-Galtsov (1989) found black-hole solutions to Einstein-Yang-Mills with non-Abelian gauge field hair; they are dynamically unstable and have not been observed.
• Boson stars and gravastars. Bound configurations of scalar or exotic matter can be black-hole impostors with non-Kerr multipole structure. EHT imaging of M87* and Sgr A* and LIGO ringdown spectra both constrain such alternatives.
• Soft hair / Bondi-Metzner-Sachs charges. Strominger and collaborators argue that horizons can carry an infinite tower of soft photon and graviton charges associated with BMS supertranslations. The proposal aims to address the information paradox without violating the classical theorem.
• Higher-dimensional hair. In D ≥ 5 the no-hair theorem does not hold; Emparan-Reall (2002) found rotating black rings with the same (M, J) as a Myers-Perry black hole but a topologically different horizon. Uniqueness fails dramatically in higher D.

Connection to the information paradox

The no-hair theorem is the classical face of the black-hole information paradox. Classically, if the exterior is described by three numbers, then everything else fell in and is lost to the outside world. That sounds benign: the infalling matter is inside the hole, and presumably its information is preserved there. The paradox arises when one couples this classical statement to quantum mechanics. Hawking's 1974 calculation showed that the hole evaporates thermally, and a thermal spectrum encodes only M (the temperature is T_H ∝ 1/M). If the evaporated spectrum is genuinely thermal, the information that distinguished the infalling matter is destroyed — violating unitary quantum mechanics.

The classical no-hair theorem is therefore not the paradox by itself; it is the classical input that creates the paradox once quantum field theory is added. The 2019-2020 "Page curve" calculations recover unitarity by showing that subtle correlations among the Hawking quanta carry the missing information out after the Page time — formally, "hair" returns at the quantum level, encoded in entanglement-wedge islands inside the hole.

Where no-hair shows up

• LIGO/Virgo/KAGRA ringdowns. The post-merger phase of every binary BH coalescence is a no-hair test. After GW150914, the LSC Testing GR papers extract one or two QNM frequencies per event and check Kerr consistency. Multi-mode "spectroscopy" of high-SNR events is the cleanest extant probe.
• Event Horizon Telescope shadow imaging. The size and shape of the shadow of M87* and Sgr A* depend only on (M, J). EHT measurements are consistent with Kerr to ~10% precision and rule out a class of horizonless alternatives.
• Extreme-mass-ratio inspirals (EMRIs). LISA (launch ~2035) will detect inspirals of compact stellar objects into supermassive black holes over months. The inspiral track is sensitive to the full multipole structure (M_ℓ + i S_ℓ) of the central object and provides the most precise Kerr-spectroscopy test currently planned.
• Astrophysical Q ≈ 0 constraint. Any charged black hole would be neutralised by selective accretion of ambient plasma within microseconds (Wald 1974). This means the astrophysical sample lives on the Kerr submanifold — Q is observationally forced to zero, not just assumed.
• Numerical relativity simulations. Every BBH merger simulation since Pretorius (2005) confirms that the remnant settles to Kerr in a few light-crossing times. Numerical waveforms are calibrated against the Kerr QNM spectrum and provide the templates LIGO uses.

Common pitfalls

• Confusing the theorem with full information loss. No-hair says the classical exterior is described by three numbers. It does not say information is destroyed in the underlying quantum theory; that is a separate (and unresolved) question.
• Forgetting Hawking's rigidity step. Carter's Kerr-uniqueness proof assumes axisymmetry. Without Hawking's rigidity theorem (1972), a stationary but not-axisymmetric black hole would not be excluded a priori.
• Treating astrophysical hair as theoretical hair. Accretion disks and magnetospheres are not hair in the sense of the theorem — they are environment. The theorem applies to the isolated stationary exterior of the hole in electrovac.
• Mistaking the theorem for assumption-free. The classical theorem assumes electrovac (no scalar fields, no Yang-Mills), no torsion, asymptotic flatness, regularity of the horizon, and four dimensions. Relaxing any of these reopens the hair question.
• Treating "tests of no-hair" as binary. LIGO does not "confirm" or "refute" no-hair; it constrains the magnitude of any deviation. The relevant figure of merit is the parameterised post-Einsteinian (ppE) bound on extra mode frequencies and damping times, currently at the 10⁻¹ - 10⁻² level.