Skip to content

Gravity as Effective Field Theory

Gravity is a predictive quantum effective field theory when every claim is tied to a low-energy observable, an operator basis, a gauge and field parametrization, explicit internal loop content, and expansion parameters below their breakdown values. The framework quantizes the massless metric degrees of freedom and includes every local interaction allowed by diffeomorphism invariance; it does not claim ultraviolet completeness.

Helpful background. Effective Field Theory of Gravity: Architecture and Power Counting supplies the generic EFT organization; Local Field Redefinitions and the Equivalence Theorem supplies basis equivalence; One-Loop Matter Effective Actions in Curved Space supplies the matter-only determinant; and Effective-Action Variation, Stress Tensors, and Consistency Checks supplies background response.

With the site’s (+---) metric and curvature convention, define the reduced Planck mass by MPl2=(8πG)1M_{\mathrm{Pl}}^2=(8\pi G)^{-1} and use

SEH=MPl22d4xg(R2Λ).S_{\mathrm{EH}} =-\frac{M_{\mathrm{Pl}}^2}{2} \int\mathrm d^4x\,\sqrt{-g}\,(R-2\Lambda).

This sign is consequential: together with Tμν=2(g)1δΓm/δgμνT_{\mu\nu}=2(\sqrt{-g})^{-1}\delta\Gamma_{\mathrm m}/\delta g^{\mu\nu} it gives Gμν+Λgμν=8πGTμνG_{\mu\nu}+\Lambda g_{\mu\nu}=8\pi G\,T_{\mu\nu}. A local gravity EFT extends it as

SEFT=d4xg[MPl22(R2Λ)+c1R2+c2RμνRμν+c3RμνρσRμνρσ+d1ΛEFT2R3+d2ΛEFT2R2R+]+Sm.\begin{aligned} S_{\mathrm{EFT}}=\int\mathrm d^4x\sqrt{-g}\Big[ &-\frac{M_{\mathrm{Pl}}^2}{2}(R-2\Lambda) +c_1R^2+c_2R_{\mu\nu}R^{\mu\nu} +c_3R_{\mu\nu\rho\sigma}R^{\mu\nu\rho\sigma}\\ &+\frac{d_1}{\Lambda_{\mathrm{EFT}}^2}R^3 +\frac{d_2}{\Lambda_{\mathrm{EFT}}^2}R\nabla^2R +\cdots\Big]+S_{\mathrm m}. \end{aligned}

In four boundary-free dimensions the Euler combination removes one curvature-squared bulk coupling, but boundaries, other dimensions, or topology-sensitive observables require its surface and global contributions. Expanding

gμν=gˉμν+2MPlhμνg_{\mu\nu}=\bar g_{\mu\nu} +\frac{2}{M_{\mathrm{Pl}}}h_{\mu\nu}

canonically normalizes the graviton up to the chosen gauge. Let R\mathcal R denote the largest independent curvature scale sampled in the observable’s physical tetrad. A typical calculation tracks

ϵp=p2(4πMPl)2,ϵR=RΛEFT2,ϵD=p2ΛEFT2,ϵspNeff(p)p2(4πMPl)2.\epsilon_p=\frac{p^2}{(4\pi M_{\mathrm{Pl}})^2}, \qquad \epsilon_R=\frac{\mathcal R}{\Lambda_{\mathrm{EFT}}^2}, \qquad \epsilon_D=\frac{p^2}{\Lambda_{\mathrm{EFT}}^2}, \qquad \epsilon_{\mathrm{sp}} \sim\frac{N_{\mathrm{eff}}(p)p^2}{(4\pi M_{\mathrm{Pl}})^2}.

Coefficient enhancements, weak-field parameters such as GM/rGM/r, and secular duration are additional observable-specific controls. At fixed order, only finitely many local coefficients enter. Massless loops also generate nonanalytic functions such as log(p2)\log(-p^2) that no local coefficient can absorb; those terms carry robust long-distance information Donoghue 1994, §§II–IV.

The structure map separates the two inputs most often confused: a matter determinant contains only internal matter lines, whereas quantum-gravity loops contain internal graviton and Faddeev–Popov ghost lines.

Gravity EFT proceeds from power counting and operator reduction through gauge-fixed graviton, ghost, and matter sectors to invariant low-energy observables

Local operators absorb short-distance dependence, while sector-resolved loops and nonanalytic terms feed matched relational, asymptotic, or curvature observables within a declared cutoff. The map is schematic and not to scale.

The chapter’s leaves follow the order in which assumptions enter a calculation.

  1. Applying EFT Power Counting to Gravity assigns momentum, curvature, loop, and species order.
  2. Curvature Operator Bases and Field Redefinitions reduces local terms without discarding physical data.
  3. Background-Field Quantization, Gauge Fixing, and Ghosts constructs the metric and ghost Hessians.
  4. Local Supersymmetry and Supergravity as Low-Energy EFT adds the gravitino and local-supersymmetry constraints.
  5. One-Loop Graviton EFT isolates graviton/ghost divergences and nonanalytic amplitudes.
  6. Matter Contributions to Gravitational Matching matches matter loops without double counting.
  7. Heavy Thresholds, Species, and the Gravitational Cutoff makes active-field and multiplicity estimates scale dependent.
  8. Matching Amplitudes to Background Observables translates common invariant data between asymptotic and curved settings.
  9. Graviton and Matter Nonlocal Form Factors keeps sector coefficients and causal prescriptions separate.
  10. Long-Distance Quantum Corrections extracts infrared tails from nonanalytic momentum dependence.
  11. Relational and Gauge-Invariant Gravitational Observables attaches predictions to physical quantities.
  12. Vacuum Energy and the Cosmological Constant separates threshold sensitivity from the measured parameter.
  13. Higher-Derivative Poles, Ghost Diagnostics, and Order Reduction classifies poles relative to the cutoff.
  14. Validity, Unitarity, and Breakdown combines the independent stopping conditions.

The following comparison is the chapter-wide contract. “Potential” always means a declared matching convention; the invariant information resides in the associated amplitude, scattering observable, or relational response.

Calculation or observableDynamical fields and internal loopsLocal basis and field redefinitionsGauge-invariant datumThreshold or nonanalytic inputExpansion and decisive failure
Tree graviton–matter amplitudeExternal matter/gravitons; no loopEH plus required contact operators; EOM-redundant terms movableOn-shell amplitudeHeavy physics in Wilson coefficientsp/ΛEFTp/\Lambda_{\mathrm{EFT}}; partial waves or omitted operators become large
Matter effective actionInternal matter only; metric prescribedCurvature basis with matter-threshold matchingRenormalized background responseMass-dependent form factors and cutsCurvature/derivative expansion fails near threshold or strong background
Graviton one-loop amplitudeInternal graviton and ghost lines, plus declared mixed diagramsFour-derivative counterterm closureOn-shell nonanalytic amplitudeMassless cuts and log(p2)\log(-p^2)p/(4πMPl)p/(4\pi M_{\mathrm{Pl}}) or species-enhanced loop reaches unity
Heavy-threshold matchingHeavy fields integrated out onceLocal Wilson coefficients below the massEquality of low-energy amplitudes/responsesSmooth decoupling around pmp\sim mExplicit heavy loop plus matched coefficient double counts
Weak-background responseFields declared in the 1PI or in-in actionBasis translated before variationSmeared curvature or relational responseRetarded continuation of nonlocal kernelsNo asymptotic S-matrix assumed; instability or curvature expansion fails
Long-distance tailMassless loop sectors resolvedAnalytic terms remain coefficient/scheme dependentScattering angle, time delay, or other invariant responseNonanalytic p\lvert p\rvert and log(p2)\log(-p^2) piecesExtrapolation to rΛEFT1r\sim\Lambda_{\mathrm{EFT}}^{-1} is forbidden
Relational observableGauge-fixed fields internally, clocks/dressings externallyInvariant under local EOM redefinitions to retained orderEvent-defined curvature, asymptotic charge, or dressed amplitudeMatching residual tested across gauges/basesBare coordinate component or incomplete clock system
Vacuum energyAll thresholds contributing to the zero-derivative termRenormalized Λ\Lambda is an independent measured couplingCurvature response after matchingContributions scale as heavy masses to the fourth powerRegulator power divergence mistaken for an observable, or naturalness claimed solved
Higher-derivative propagatorEH pole plus perturbative insertions below cutoffEOM-redundant pieces removed only within observablePole/residue of a controlled low-energy amplitudeExtra exact pole compared with ΛEFT\Lambda_{\mathrm{EFT}}Above-cutoff root treated as an EFT particle or order reduction used across a genuine low pole

The failure map is therefore diagnostic rather than universal: loop misclassification, redundant operators, gauge dependence, spurious poles, and missing truncation errors require different repairs.

Matter-loop and graviton-loop confusion, redundant operators, gauge-dependent outputs, spurious higher-derivative poles, and omitted errors trigger distinct failures

A low-energy gravitational claim survives only when its fields, basis, gauge-invariant observable, loop sector, threshold treatment, and truncation uncertainty are simultaneously controlled. The map is schematic and not to scale.

A result should leave this chapter with the action and curvature signs; canonical normalization; active degrees of freedom; internal graviton, ghost, matter, and mixed lines; renormalization scheme and scale; field-redefinition class; gauge-invariant observable; analytic and nonanalytic pieces; matching conditions; cutoff; and numerical remainder estimate. Pure-gravity one-loop on-shell finiteness in a special theory is not off-shell finiteness, matter-coupled finiteness, or ultraviolet renormalizability. Likewise, a species scale is a regime-dependent estimate and a higher-derivative pole above the cutoff is not an additional EFT state.

Generic matching theory remains in Renormalization and EFT. Matter-only determinants remain in Chapter 8. Detailed ultraviolet completions and nonperturbative quantum geometry remain with Holography and Quantum Gravity.

  • Burgess, C. P. “Quantum Gravity in Everyday Life: General Relativity as an Effective Field Theory.” Living Reviews in Relativity 7, 5 (2004). doi:10.12942/lrr-2004-5. Open PDF
  • Donoghue, J. F. “General Relativity as an Effective Field Theory: The Leading Quantum Corrections.” Physical Review D 50, 3874–3888 (1994). doi:10.1103/PhysRevD.50.3874. Open PDF