General Relativity as a Quantum Effective Field Theory

General relativity is a predictive quantum effective field theory at energies and curvatures below its cutoff. The action contains every local generally covariant operator, while power counting determines how many coefficients enter at a desired accuracy. Unknown ultraviolet physics changes local Wilson coefficients; massless propagation fixes nonanalytic long-distance terms. This chapter turns that distinction into a calculation method and states where it stops being controlled.

Helpful background. Flat-Spacetime Quantum Field Theory supplies matching, running, and amplitudes; Einstein Equations supplies the leading dynamics; Quantum Information supplies unitarity language; and Quantum Gravity supplies the scale, regime, and evidence conventions.

Predictivity is order by order

At momentum $E$ well below a heavy scale $\Lambda$, write the Lagrangian as a derivative expansion,

$$\mathcal L_{\mathrm{EFT}}=\sqrt{-g}\left[ \frac{R-2\Lambda_{\mathrm{cc}}}{16\pi G} +c_1R^2+c_2R_{\mu\nu}R^{\mu\nu} +\frac{d_1}{\Lambda^2}R^3+\cdots\right].$$

Only finitely many terms contribute through a fixed power of $E/\Lambda$ and $GE^2$. Loops renormalize the same operator basis. A new divergence therefore signals the coefficient needed at the next order, not a loss of every lower-order prediction. Donoghue gives the low-energy construction and its separation of local from long-distance effects Donoghue 2023, §§2–6.

The observables here are on-shell amplitudes and quantities extracted from them with declared conventions. Dimensional regularization and covariant gauge fixing organize intermediate expressions. Physical results must be gauge independent; local coefficients depend on renormalization scheme until fixed by matching or measurement.

Six questions form the EFT workflow

1. The Gravitational EFT Action classifies generally covariant operators and Wilson coefficients.
2. Covariant Power Counting compares loop, curvature, and higher-derivative orders.
3. One-Loop Renormalization of Gravity separates the special pure-gravity on-shell cancellation from generic counterterm requirements.
4. Quantum Corrections to Long-Range Potentials extracts nonanalytic momentum terms and their power-law spatial tails.
5. Classical Physics from Quantum Amplitudes explains why loops can contain classical post-Minkowskian contributions.
6. Breakdown, Matching, and UV Sensitivity identifies the first uncontrolled ratio and maps ultraviolet models only onto the coefficients they actually determine.

The recurring discipline is to report the light field content, state or external particles, kinematics, operator basis, regulator, renormalization scale, boundary conditions, retained order, and first omitted correction.

Results and open questions as of 2026-08-20

The low-energy framework is established; extrapolation beyond its cutoff remains unresolved.

Result or claim	Evidence or calculation	Scope and limitation
Covariant operator expansion	Locality and diffeomorphism invariance organize all counterterms by derivatives	Assumes a local low-energy description and a separation of scales
One-loop renormalization	Curvature-squared divergences are controlled; pure four-dimensional gravity has a special on-shell cancellation ’t Hooft and Veltman 1974, §§3–6	Matter, off-shell quantities, and higher loops require additional coefficients
Long-distance predictions	Massless cuts fix logarithms and square-root momentum terms; local UV physics cannot change them Donoghue 1994, §§II–V	Coefficients depend on the complete light spectrum and observable definition
UV matching	A specified heavy theory determines local Wilson coefficients by equating low-energy observables	Matching does not validate the heavy theory empirically or extend EFT beyond its control scale

The conservative position says EFT is deliberately agnostic about microscopic gravity and should make only low-energy claims. A permissive position treats patterns in measured or calculated Wilson coefficients as clues to ultraviolet structure. These competing interpretations agree that nonanalytic infrared terms are more universal than subtraction polynomials, and that the ultraviolet completion remains an open question.

The conclusions could change when a new matching calculation changes a Wilson-coefficient relation, an independent loop analysis changes a universal nonanalytic term, a light degree of freedom alters the EFT spectrum, or evidence after 2026-08-20 changes the cutoff or interpretation.

Boundaries and correspondence

The expansion is controlled when every invariant momentum and curvature scale is small compared with its suppression scale and when large logarithms or long-time secular terms are absent or resummed. It can fail below the Planck energy if a new light threshold, strong collective effect, or large occupation invalidates the chosen variables. Conversely, a macroscopic mass does not cause failure when local curvature and transfer momenta remain small.

Higher-derivative operators should normally be treated perturbatively. Resumming a truncated polynomial propagator can manufacture extra poles near the cutoff, where omitted operators are equally important. Such poles are not reliable new particles without a matched ultraviolet theory.

Review the chapter

Diagnostic. With the mostly-plus metric and principal branch, a low-transfer amplitude has $\mathcal M(q^2)=a+bq^2+cq^2\log[(q^2-i0)/\mu^2]$, where $\mu$ is a reference mass. At this EFT order, with the lower-order light-field couplings fixed, which coefficients can local counterterms change, and which term gives a long-distance tail?

Solution

Local contact terms expand as polynomials in momentum, so matching and scheme choices can shift $a$ and $b$. They cannot remove the logarithmic branch point. Thus $c$ follows from the low-energy spectrum and lower-order couplings at this order, and produces a power-law tail rather than a contact interaction. Higher-order loops with matched vertices can depend on further Wilson coefficients; the universality claim always has a specified EFT order.

References

• ’t Hooft, Gerard, and Martinus Veltman. “One-Loop Divergencies in the Theory of Gravitation.” Annales de l’Institut Henri Poincaré A 20 (1974): 69–94. Open scan.
• Donoghue, John F. “General Relativity as an Effective Field Theory: The Leading Quantum Corrections.” Physical Review D 50 (1994): 3874–3888. DOI.
• Donoghue, John F. “Quantum General Relativity and Effective Field Theory.” 2023. arXiv:2211.09902.