The hierarchy problem

The hierarchy problem is the sharpest discomfort in the Standard Model, and it is a problem of arithmetic before it is a problem of physics. The theory works: its gauge sector, built on $SU(3)\times SU(2)\times U(1)$ and fixed almost entirely by symmetry, predicted the pattern of weak couplings that LEP and SLC confirmed to a fraction of a percent. The trouble lives one line down in the Lagrangian, in the Higgs sector, where a single dimensionful parameter — the mass term $-\mu^2\,\phi^\dagger\phi$ that triggers electroweak symmetry breaking — sets the scale $v\approx246$ GeV of everything we call "low energy". Nothing in the theory protects that number. Quantum corrections drag a scalar mass toward the highest energy the theory knows about, and if that scale is the Planck mass $M_{\rm Pl}\approx1.2\times10^{19}$ GeV, the correction is seventeen orders of magnitude too large. To keep the Higgs at 125 GeV against a Planck-sized correction, the bare parameter and the loop correction must cancel to some thirty decimal places — a cancellation the theory neither predicts nor forbids, but simply requires as an input. That extreme, unexplained sensitivity is the hierarchy problem. It is not a sign the Standard Model is wrong; it reproduces data beautifully. It is a sign the theory is incomplete — that some deeper structure (supersymmetry, a composite Higgs, extra dimensions) should render the electroweak scale natural rather than accidental. This article makes the sensitivity quantitative, traces its origin, and surveys the proposed answers.

Key takeaways

The electroweak scale $v\approx246$ GeV sits $\sim10^{17}$ times below the Planck scale, and nothing in the Standard Model explains the gap — the big hierarchy.
A scalar mass has no symmetry protecting it, so radiative corrections scale as $\delta m_h^2\sim\Lambda^2$ — quadratically sensitive to the highest scale $\Lambda$ in the theory.
With $\Lambda\sim M_{\rm Pl}$ the correction is $\sim10^{33}$ times $m_h^2$ , so the bare mass must cancel it to $\sim30$ decimal places: an unnatural fine-tuning.
't Hooft naturalness says a small parameter is only natural if a symmetry protects it — fermion masses (chiral) and gauge masses (gauge) qualify; the Higgs mass does not.
Supersymmetry cancels the quadratic divergence boson-against-fermion; other cures lower the fundamental cutoff (compositeness, extra dimensions).
Key idea: because the Higgs mass is a scalar and no symmetry forbids it, quantum corrections make it quadratically sensitive to the highest energy scale — so the observed $125$ GeV, sitting far below the Planck scale, demands either an enormous fine-tuning or new physics near the TeV that stabilises it.

The core insight

The Higgs mass is the only Standard Model parameter with no symmetry protecting it, so quantum corrections drive it quadratically toward the highest scale in the theory; keeping it at $125$ GeV far below $M_{\rm Pl}$ therefore requires either a $\sim30$ -decimal fine-tuning or new physics near the TeV scale.

Where this lives today

The hierarchy problem is the single strongest motivation for physics beyond the Standard Model, and the LHC was built in large part to test its favoured answer. Naturalness predicted new particles near the TeV — top squarks, gluinos, and the rest of a supersymmetric spectrum — light enough to stabilise the Higgs mass. Two decades of searches at the Tevatron and LHC found none, pushing coloured superpartners above $\sim1$ – $2$ TeV and reintroducing tuning at the percent level: the little hierarchy problem. Composite-Higgs and "little Higgs" models make the Higgs a pseudo-Nambu–Goldstone boson protected until a compositeness scale of a few TeV, probed through deviations in Higgs couplings and searches for vector-like top partners. Warped and large extra dimensions (Randall–Sundrum; Arkani-Hamed–Dimopoulos–Dvali) attack the problem at its root by lowering the fundamental gravitational scale to the TeV, so there is no desert to tune across. Newer ideas — the cosmological relaxion and neutral-naturalness scenarios — try to decouple the stabilising physics from coloured particles the LHC would already have seen. A future $100$ -TeV collider (FCC-hh) is the primary proposal to settle whether naturalness is realised near the TeV or the electroweak scale is simply tuned.

The physical picture

Two pictures carry the whole problem. The first is the raw disparity of scales: the Higgs vacuum expectation value $v\approx246$ GeV sits an astonishing seventeen decades below the scale where gravity becomes strong. A scalar mass at the bottom of that ladder has no reason to stay there — quantum mechanics wants to lift it to the top.

Figure

The electroweak scale sits about 17 orders of magnitude below the Planck scale — v ≈ 246 GeV and m_h ≈ 125 GeV against M_Pl ≈ 1.2×10¹⁹ GeV, a ratio v/M_Pl ≈ 2×10⁻¹⁷ that the Standard Model leaves unexplained

The second picture is the mechanism that makes the gap fragile. The physical Higgs mass is the bare Lagrangian parameter plus quantum corrections from every particle the Higgs couples to. The strongest coupling is to the top quark, whose loop drags the mass-squared down by an amount growing as the cutoff squared. If a heavy scalar exists — a superpartner of the top — its loop pushes back up with the opposite sign, and if supersymmetry fixes the couplings just so, the two enormous contributions cancel decade by decade, leaving only a mild logarithm. That cancellation, or its absence, is the entire story.

Figure

One-loop corrections to the Higgs mass squared: a top-quark loop pulls it down by roughly −(3y_t²/8π²)Λ² while each scalar-partner loop pushes up by +(λ_S/16π²)Λ² — when supersymmetry fixes λ_S = y_t² the quadratic pieces cancel and only a mild ln Λ term survives

The core equations

Electroweak symmetry breaking is driven by a single potential with two parameters. The mass term has the "wrong" sign, $-\mu^2$ , so the field rolls off the origin to a minimum at nonzero value; the quartic $\lambda$ stabilises it. Minimising fixes the vacuum expectation value and the physical Higgs mass.

V(\phi)=-\mu^2\,\phi^\dagger\phi+\lambda\,(\phi^\dagger\phi)^2,\qquad v=\frac{\mu}{\sqrt\lambda}\approx246\ \text{GeV},\qquad m_h^2=2\mu^2=2\lambda v^2 \qquad (1)

where: $\phi$ — the complex Higgs doublet; $\mu^2>0$ — the mass parameter at the heart of the problem; $\lambda\approx0.13$ — the quartic self-coupling fixed by the measured $m_h$ ; $v\approx246$ GeV — the vacuum expectation value; $m_h\approx125$ GeV — the physical Higgs mass. Every low-energy mass — $W$ , $Z$ , and every fermion — is proportional to $v$ , so $\mu$ sets the entire electroweak scale. The $\mu^2$ in the Lagrangian is bare — not directly measurable. What experiment fixes is the physical mass: the bare value plus every quantum correction, evaluated with a cutoff $\Lambda$ standing for the scale where new physics enters. Renormalisation says the two $\Lambda$ -dependences must conspire to leave the sum finite and equal to the measured $m_h$ .

m_h^2\big|_{\text{phys}}=m_{h,0}^2(\Lambda)+\delta m_h^2(\Lambda) \qquad (2)

where: $m_h^2|_{\text{phys}}=(125\ \text{GeV})^2$ — the measured value; $m_{h,0}^2(\Lambda)$ — the bare mass-squared, a free input; $\delta m_h^2(\Lambda)$ — the summed loop correction. Nothing is wrong with this equation; the problem is how large the two right-hand terms must be relative to their sum. The dominant correction comes from the top quark, which has by far the largest coupling to the Higgs. A fermion loop contributes with a negative sign and grows as the square of the cutoff — the quadratic divergence a scalar suffers and a fermion or gauge mass does not.

\delta m_h^2\big|_{\text{top}}=-\frac{3\,y_t^2}{8\pi^2}\,\Lambda^2+(\text{logarithms}),\qquad y_t=\frac{\sqrt2\,m_t}{v}\approx0.99 \qquad (3)

where: $y_t$ — the top Yukawa coupling, close to one; the factor $3$ — the number of quark colours; $\Lambda$ — the cutoff, the scale of first new physics. The coefficient $3y_t^2/8\pi^2\approx0.037$ is of order a percent, but times $\Lambda^2$ it dwarfs $m_h^2$ unless $\Lambda$ is near the TeV. Summing every Standard Model particle gives the full one-loop coefficient of $\Lambda^2$ — the Veltman expression. The top's large negative contribution wins, so the coefficient does not vanish; it would cancel only for an unphysical Higgs mass.

\delta m_h^2=\frac{3\,\Lambda^2}{8\pi^2 v^2}\left(m_h^2+m_Z^2+2m_W^2-4m_t^2\right) \qquad (4)

where: the bracket collects the Higgs, $Z$ , $W$ , and top masses with multiplicities from their spins and couplings; numerically it is $\approx-(287\ \text{GeV})^2$ , negative and top-dominated. The Veltman condition — that the bracket vanish — would need $m_h\approx313$ GeV, which is excluded, so no accidental Standard-Model cancellation saves us. The discomfort is made precise by a fine-tuning measure: how sensitive $m_h^2$ is to a fractional change in each input $a_i$ . It is of order one for a natural theory and enormous for a fine-tuned one.

\Delta=\max_i\left|\frac{\partial\ln m_h^2}{\partial\ln a_i}\right|\sim\frac{|\delta m_h^2|}{m_h^2}\sim\frac{\Lambda^2}{m_h^2} \qquad (5)

where: $\Delta$ — the Barbieri–Giudice fine-tuning parameter; $\Delta=1$ means no tuning, $\Delta=100$ means percent-level stability. For $\Lambda\sim M_{\rm Pl}$ , equation (5) gives $\Delta\sim10^{33}$ — the headline number of the whole subject.

Derivation AWhy a scalar mass is quadratically sensitive to the cutoff

The goal here is to see exactly why the Higgs mass alone among Standard Model parameters is dragged toward the highest energy scale in the theory. We trace the quadratic divergence to its source in a loop integral and contrast it with the merely logarithmic running of fermion and gauge masses, seeing that the difference is not an accident of calculation but a consequence of which symmetries forbid which mass terms. The punchline: a scalar mass is unprotected, so its correction knows about the full ultraviolet.

Step 1. Compute the top-quark loop correcting the Higgs two-point function. The Yukawa vertex $-y_t\,\bar t\,t\,h$ appears twice, giving $\delta m_h^2\sim -y_t^2\int^\Lambda \frac{d^4k}{(2\pi)^4}\,\frac{k^2}{(k^2-m_t^2)^2}$ . The $k^2$ from the fermion trace against four denominator powers leaves the integrand $\sim1/k^2$ at large $k$ , so $\int d^4k/k^2\sim\Lambda^2$ . Why: a fermion loop carries an extra power of loop momentum in its numerator, and that single power turns a would-be logarithm into a quadratic divergence for the scalar it corrects.

Step 2. Read off sign and coefficient: the closed fermion loop supplies a minus sign, three colours multiply it, and the reduction gives $\delta m_h^2=-\tfrac{3y_t^2}{8\pi^2}\Lambda^2$ plus terms growing only as $\ln\Lambda$ . Why: the negative sign is exactly what lets a bosonic partner (positive contribution) cancel it — the seed of the supersymmetric solution.

Step 3. Contrast with a fermion mass. Set a fermion's mass to zero and the Lagrangian gains a chiral symmetry $\psi\to e^{i\gamma_5\alpha}\psi$ that forbids the mass term; a nonzero mass therefore breaks that symmetry, and any correction must be proportional to the mass itself, $\delta m_f\propto m_f\ln\Lambda$ . The divergence is only logarithmic and multiplicative. Why: a symmetry that appears when a parameter vanishes forces every correction to that parameter to be proportional to it — the parameter cannot be regenerated large from zero.

Step 4. Now ask what symmetry protects the scalar mass. Setting $\mu^2\to0$ enhances no symmetry of the Standard Model — a massless scalar is not special, unlike a massless fermion (chiral) or gauge boson (gauge invariance). With nothing forbidding the mass term, corrections are additive and drink from the whole ultraviolet, $\delta m_h^2\sim\Lambda^2$ . Why: this missing symmetry is the technical statement of the hierarchy problem — the scalar mass is the one dimensionful parameter quantum mechanics is free to make as large as the cutoff allows.

Go deeper · Derivation B · 1 minThe size of the fine-tuning at the Planck scale

Derivation BThe size of the fine-tuning at the Planck scale

The aim is to convert the abstract worry "the Higgs mass is sensitive to the cutoff" into a single, visceral number: how many decimal places of cancellation the Standard Model demands if the next scale of physics is gravitational. We take the cutoff to the Planck mass, evaluate the top-loop correction, and compare it to the measured Higgs mass-squared — the reason the word "unnatural" attaches to this problem rather than merely "curious".

Step 1. Fix the cutoff at $\Lambda=M_{\rm Pl}=1.22\times10^{19}$ GeV, where gravitational quantum effects can no longer be ignored, and evaluate the top correction of equation (3): $\delta m_h^2=\tfrac{3y_t^2}{8\pi^2}\Lambda^2=0.037\times(1.22\times10^{19})^2\ \text{GeV}^2\approx5.6\times10^{36}\ \text{GeV}^2$ . Why: the Planck scale is the most conservative cutoff — gravity guarantees new physics no later than here — so this is the smallest the problem can be if nothing intervenes below.

Step 2. Compare with the measured value $m_h^2=(125\ \text{GeV})^2\approx1.57\times10^4\ \text{GeV}^2$ : the ratio is $m_h^2/\delta m_h^2\approx2.8\times10^{-33}$ . Why: this ratio is the fraction to which the bare mass and the loop correction must agree — how precisely two big numbers have to cancel.

Step 3. Interpret via equation (2): the bare $m_{h,0}^2$ and the correction $\delta m_h^2$ are each $\sim10^{36}$ GeV², yet their sum is $\sim10^{4}$ GeV², so the two independent inputs must share their leading $\sim33$ digits and differ only in the $34$ th. Why: a $\sim30$ -decimal cancellation between quantities with no known relation is what physicists mean by an unnatural fine-tuning — logically allowed, but a conspiracy demanding explanation.

Step 4. Note the sensitivity to the cutoff: if the first new physics sits at the grand-unification scale $\Lambda\sim10^{16}$ GeV the tuning eases to $\sim1$ part in $10^{26}$ , and if it sits at a few TeV the correction is comparable to $m_h^2$ and the tuning vanishes. Why: the tuning is monotonic in how far away new physics lies, so "solve the hierarchy problem" means "put new physics near the TeV scale".

Go deeper · Derivation C · 2 minHow supersymmetry cancels the quadratic divergence

Derivation CHow supersymmetry cancels the quadratic divergence

The goal is to show precisely how a symmetry relating bosons and fermions removes the quadratic divergence, and why that success has a price the LHC keeps raising. We add a scalar partner to the top, compute its loop, exploit the opposite sign of bosonic and fermionic loops, and impose the coupling relation supersymmetry enforces. We then see that once the partner is heavier than the top — as it must be — a residual correction proportional to the partner mass returns, reopening the problem in milder form.

Step 1. Introduce a complex scalar $S$ (a "stop") coupling to the Higgs through $-\lambda_S|H|^2|S|^2$ . Its loop gives $\delta m_h^2=+\tfrac{\lambda_S}{16\pi^2}\Lambda^2$ — same quadratic growth as the top loop, but positive, since a bosonic loop lacks the minus sign of a closed fermion loop. Why: the sign difference between boson and fermion loops is the single fact that makes cancellation possible.

Step 2. Count states and impose the supersymmetric relation. Each top has two real scalar partners (from $\tilde t_L$ and $\tilde t_R$ ), and unbroken supersymmetry fixes $\lambda_S=y_t^2$ . Adding the top loop of equation (3) to the two scalar loops gives $-\tfrac{y_t^2}{8\pi^2}\Lambda^2+2\cdot\tfrac{y_t^2}{16\pi^2}\Lambda^2=0$ : the quadratic divergences cancel exactly. Why: supersymmetry is precisely the structure that ties the scalar coupling to the fermion Yukawa, so the cancellation is automatic, order by order, rather than tuned.

Step 3. Break the symmetry, as nature must, giving the scalar a mass $m_{\tilde t}>m_t$ . The leftover correction is now only logarithmically divergent, $\delta m_h^2\simeq\tfrac{3y_t^2}{8\pi^2}\,(m_{\tilde t}^2-m_t^2)\,\ln(\Lambda/m_{\tilde t})$ : the dangerous $\Lambda^2$ is gone, and the residual is set by the superpartner mass, not $M_{\rm Pl}$ . Why: the cutoff sensitivity is traded for sensitivity to a new scale that can be near the TeV, turning a $10^{33}$ tuning into a mild one.

Step 4. Read the naturalness bound. For the residual not to exceed $m_h^2$ by a large factor, the stops must be light — a few hundred GeV to about a TeV. LHC searches now exclude stops below roughly $1.1$ TeV, so the residual again exceeds $m_h^2$ by one-to-two orders of magnitude. Why: this reopened, milder tuning of $\sim1$ – $10\%$ is the "little hierarchy problem" — the reason naturalness is under pressure even though the quadratic divergence is solved.

Limiting cases

Case	Result	Meaning
New physics near the TeV, $\Lambda\sim1$ TeV	$\delta m_h^2\sim m_h^2$ , $\Delta\sim\mathcal{O}(1)$	natural — the correction is the size of the answer; nothing to tune
Cutoff at grand unification, $\Lambda\sim10^{16}$ GeV	$\Delta\sim10^{26}$	tuning to $\sim26$ decimals
Cutoff at the Planck scale, $\Lambda\sim10^{19}$ GeV	$\Delta\sim10^{33}$	the maximal, most-quoted fine-tuning
Exact supersymmetry, $m_{\tilde t}=m_t$	$\delta m_h^2=0$ (quadratic and log)	boson and fermion loops cancel completely
Broken supersymmetry, $m_{\tilde t}>m_t$	$\delta m_h^2\sim\tfrac{y_t^2}{8\pi^2}m_{\tilde t}^2\ln\Lambda$	residual set by the superpartner scale, not $\Lambda$
Dimensional regularisation	no explicit $\Lambda^2$ term	sensitivity hides in heavy-particle thresholds $M^2$ , not in the regulator

Common trapthe misconception: "dimensional regularisation makes the quadratic divergence vanish, so there is no real problem"

True, in dimensional regularisation the $\Lambda^2$ piece never appears — the scheme sets scaleless power divergences to zero, and the Higgs mass runs only logarithmically. But this is bookkeeping, not physics. The genuine content of the problem is physical: whenever a heavy particle of mass $M$ couples to the Higgs — a grand-unified partner, a heavy right-handed neutrino, any state in the desert — its threshold contributes $\delta m_h^2\sim\tfrac{y^2}{16\pi^2}M^2$ , a real, scheme-independent correction proportional to the heavy physical mass, not to any regulator. Dimensional regularisation merely relabels the cutoff by the mass of the next real thing, and the fine-tuning reappears intact the moment you admit any high scale. No choice of regulator can make heavy particles stop existing. (A related confusion: that the tuning means the Standard Model is inconsistent — it is not; it is consistent and predictive, just unnatural.)

Worked examples

1 · How fine-tuned is the Standard Model with a Planck cutoff? (fully worked)

Take the top-quark loop as the dominant correction and assume no new physics below the Planck scale. Quantify the cancellation the Higgs mass demands.

Step 1. Evaluate the correction coefficient. With $y_t=\sqrt2\,m_t/v=\sqrt2\times172.7/246.2\approx0.99$ , the coefficient is $3y_t^2/8\pi^2=3(0.99)^2/(8\times9.87)\approx\mathbf{0.037}$ — about a percent, small until multiplied by a squared Planck mass.

Step 2. Insert the cutoff $\Lambda=M_{\rm Pl}=1.22\times10^{19}$ GeV: $\delta m_h^2=0.037\times(1.22\times10^{19})^2\approx0.037\times1.49\times10^{38}\approx\mathbf{5.6\times10^{36}\ GeV^2}$ .

Step 3. Compare to the physical value $m_h^2=(125.25)^2\approx1.57\times10^4$ GeV²: the ratio is $m_h^2/\delta m_h^2\approx\mathbf{2.8\times10^{-33}}$ .

Step 4. Interpret. The bare $m_{h,0}^2$ and the loop correction are each $\sim10^{36}$ GeV² yet must sum to $\sim10^{4}$ GeV²: they must agree to about 33 decimal places. Equivalently $\Delta\approx\mathbf{3.6\times10^{32}}$ . The Standard Model with a Planck cutoff is tuned to roughly one part in $10^{33}$ — the sharpest form of the hierarchy problem.

2 · The naturalness bound on the top squark

Supersymmetry removes the $\Lambda^2$ term, leaving $\delta m_h^2\simeq\tfrac{3y_t^2}{8\pi^2}\,m_{\tilde t}^2\,\ln(\Lambda/m_{\tilde t})$ . Ask how heavy the stops may be before tuning returns, and compare with LHC limits.

Step 1. Set a tolerance: require $|\delta m_h^2|\lesssim5\,m_h^2\approx5\times1.57\times10^4\approx\mathbf{7.9\times10^{4}\ GeV^2}$ — roughly " $10\%$ natural", $\Delta\lesssim10$ .

Step 2. With a modest logarithm $\ln(\Lambda/m_{\tilde t})\approx3$ (new physics a decade above the stops), $0.037\times3\,m_{\tilde t}^2\lesssim7.9\times10^4$ gives $m_{\tilde t}\lesssim\mathbf{590\ GeV}$ ; with a high mediation scale $\ln(M_{\rm Pl}/m_{\tilde t})\approx37$ the large logarithm tightens this to $m_{\tilde t}\lesssim\mathbf{170\ GeV}$ .

Step 3. Confront experiment. LHC searches exclude a top squark below about $\mathbf{1.1\ TeV}$ over most of parameter space. Since $1.1$ TeV lies well above the $170$ – $590$ GeV natural window, the residual exceeds $m_h^2$ by one-to-two orders of magnitude. Supersymmetry solves the big hierarchy problem but, given current limits, leaves a "little" tuning of $\sim1$ – $10\%$ — the observational pressure on naturalness.

Physics of the portrait

The hierarchy problem has no single discoverer; it crystallised out of one idea — naturalness — carried by Wilson's renormalisation, sharpened by 't Hooft into a symmetry criterion, and turned on the Higgs by Susskind and Weinberg.

Origins · 1 minnaturalness and how this was discovered

naturalness

the criterion that a small dimensionless parameter is natural only if a symmetry appears when it is set to zero · Wilsonian effective field theory and the Higgs mass

Naturalness grew out of Kenneth Wilson's reformulation of quantum field theory in the early 1970s, in which a theory is an effective description valid below a cutoff and every coupling allowed by symmetry should appear with a coefficient of order one — so a scalar mass far below the cutoff is a glaring exception. Gerard 't Hooft, in his 1979 Cargèse lectures, made this a precise criterion: a small parameter is "technically natural" only if setting it to zero enlarges the theory's symmetry, so corrections to it stay proportional to the parameter. Fermion masses pass (a vanishing mass restores chiral symmetry) and gauge masses pass (gauge invariance protects them); the Higgs mass fails, because no symmetry returns when $\mu^2\to0$ . That same year Leonard Susskind, crediting Wilson, and Steven Weinberg independently applied the argument to the electroweak Higgs, noting its quadratic instability and proposing new strong dynamics (technicolor) to avoid an elementary scalar. Martinus Veltman soon wrote the full one-loop condition, and supersymmetry emerged as the symmetry that could tame it. Naturalness has since organised four decades of model building, and its confrontation with the LHC's null results is one of the defining tensions of contemporary particle physics.

History

1976

— Gildener points out the fine-tuning of scalar masses in grand unified theories

1979

— Susskind and Weinberg identify the Higgs mass instability and motivate technicolor

1979

— 't Hooft's Cargese lectures state the technical-naturalness criterion

1981

— Veltman writes the full one-loop quadratic-divergence condition; supersymmetry emerges as a cure

2012

— the LHC discovers a 125 GeV Higgs, with no superpartners in sight

Going further

Texts. Peskin & Schroeder, An Introduction to Quantum Field Theory, §12.3 (renormalisation and the cutoff) and the Higgs sections of Ch. 20 · Peskin, Beyond the Standard Model, SLAC-PUB-7479 / hep-ph/9705479, §2.1–2.2 · Schwartz, Quantum Field Theory and the Standard Model, Ch. 22–23 (naturalness and effective field theory) · Thomson, Modern Particle Physics, Ch. 17 (Higgs and beyond) · Martin, A Supersymmetry Primer, hep-ph/9709356, §1 (the hierarchy problem and its supersymmetric solution). The key numerical checks underlying this article are: (1) $m_h^2=2\lambda v^2$ gives $\lambda\approx0.13$ for $m_h=125$ GeV, $v=246$ GeV; (2) the top-loop coefficient $3y_t^2/8\pi^2\approx0.037$ with $y_t\approx0.99$ ; (3) the Planck-cutoff correction $\delta m_h^2\approx5.6\times10^{36}$ GeV² and tuning $\sim10^{-33}$ ; (4) the Veltman bracket $m_h^2+m_Z^2+2m_W^2-4m_t^2<0$ , vanishing at $m_h\approx313$ GeV; (5) the supersymmetric cancellation $-y_t^2\Lambda^2/8\pi^2+2\lambda_S\Lambda^2/16\pi^2=0$ when $\lambda_S=y_t^2$ .

Prerequisites. The Higgs mechanism · Electroweak unification · Spontaneous symmetry breaking · Standard Model summary. Leads to. Supersymmetry · Grand unification · Open questions · Extra dimensions.

Ask the tutor: Ask me to walk through the top-loop calculation line by line, or to compare how supersymmetry, a composite Higgs, and extra dimensions each make the electroweak scale natural.

Is the hierarchy problem a sign that the Standard Model is wrong?

No. The Standard Model is one of the most successful theories in the history of science, verified to percent-level accuracy, and nothing about the hierarchy problem contradicts a single measurement. The theory is fully consistent and predictive with a fine-tuned Higgs mass — write down the bare parameter that gives $125$ GeV and everything works. The problem is one of naturalness, not consistency: that bare parameter must be chosen to $\sim30$ decimal places for no reason the theory supplies. Physicists treat such an unexplained conspiracy as a clue that a deeper theory exists, much as the fine structure of atomic spectra once pointed toward quantum mechanics. It is a heuristic problem, not a logical flaw.

Why doesn't the same problem afflict the electron or the photon mass?

Because both are protected by symmetries the Higgs mass is not. Set the electron mass to zero and the Lagrangian acquires a chiral symmetry under which left- and right-handed electrons rotate by opposite phases; that symmetry forbids a mass term, so any correction to the electron mass is proportional to the mass itself and grows only logarithmically with the cutoff. The photon is protected even more strongly by gauge invariance, which forbids a photon mass outright. A scalar mass has no such shield: setting $\mu^2\to0$ restores no symmetry, so corrections are additive and quadratic in the cutoff. This is why the hierarchy problem is specific to fundamental scalars, and why the Higgs — the Standard Model's only elementary scalar — is uniquely afflicted.

What exactly does "natural" mean, and who defined it?

Naturalness is the requirement that dimensionless ratios of parameters be of order one unless a symmetry explains their smallness. Gerard 't Hooft gave the sharp version in 1979: a small parameter is "technically natural" if setting it to zero enlarges the symmetry of the theory, so that corrections to it stay proportional to the parameter. By this test a tiny fermion mass is natural (chiral symmetry protects it) but a Higgs mass far below the cutoff is not. A complementary quantitative measure, due to Barbieri and Giudice, tracks how sensitively an output depends on each input; large sensitivity signals fine-tuning. Naturalness is a guiding principle, not a law of nature — the universe is under no obligation to obey it — but it has repeatedly and correctly pointed toward new physics in the past.

Does supersymmetry completely solve the problem?

It solves the severe part and softens the rest. Exact supersymmetry pairs every fermion with a boson whose loop cancels the fermion's quadratic divergence, so the $\Lambda^2$ sensitivity vanishes and the Higgs mass is no longer dragged toward the Planck scale. But supersymmetry must be broken, since we see no mass-degenerate superpartners, and once the superpartners are heavier a residual correction returns, now proportional to the superpartner masses rather than to $\Lambda$ . Keeping that residual comparable to $m_h^2$ needs light top squarks and gluinos, ideally below a TeV — and LHC searches have excluded much of that window, leaving a mild "little" tuning of order a percent. Supersymmetry converts a $10^{33}$ problem into a $10^{2}$ one: a spectacular improvement, but not, at present, a perfectly natural theory.

Could there simply be no new physics, and the tuning be an accident we must accept?

That is a live and respectable possibility. If a multiverse realises many values of the Higgs mass parameter, anthropic selection could pick out the small value we observe: only where the electroweak scale is far below the Planck scale do complex nuclei, chemistry, and observers form. Weinberg used exactly this logic to anticipate the small cosmological constant, another severe fine-tuning, and some regard the two puzzles as linked. In this view the tuning is real but explained by selection rather than symmetry, and one should not expect TeV-scale superpartners. The cost is a loss of predictivity, which is why most physicists treat anthropic explanations as a last resort — but the LHC's null results have moved the idea from the fringe toward the mainstream of the debate.

How is the hierarchy problem related to Peskin's "arbitrary parameters" complaint?

They are two faces of the same discomfort with the Higgs sector. Peskin frames it structurally: the gauge sector is fixed almost entirely by symmetry and works beautifully, whereas the Higgs and Yukawa sector carries a crowd of undetermined parameters — $\mu^2$ , $\lambda$ , and three complex Yukawa matrices — that symmetry barely constrains. The hierarchy problem is the most acute member of that family: the one parameter, $\mu^2$ , that sets the overall electroweak scale is not merely undetermined but radiatively unstable. Both point to the conclusion Peskin draws — that the Higgs is unlikely to be a simple elementary object, and that a deeper theory should determine or replace these parameters. The arbitrariness is the symptom; the quadratic instability is where it hurts.