---
title: 'Berry–Esseen Bounds: Rates and Extensions'
url: https://www.emergentmind.com/topics/berry-esseen-bounds
type: topic
---

# Berry–Esseen Bounds: Rates and Extensions

A Berry–Esseen bound provides a quantitative measure of how quickly the distribution of a suitably normalized sum of random variables converges to a limiting law (typically Gaussian) in central limit theorems, quantifying this proximity via explicit, often optimal, rates in terms of moments and structural dependence. The Berry–Esseen framework has evolved from classical i.i.d. settings to encompass dependencies, functional statistics (e.g., U-statistics), operator-valued/complex random variables, multivariate and high-dimensional statistics, self-normalized forms, and non-asymptotic local limit theorems. New methods such as entropy-based arguments, concentration inequalities, exchangeable pairs, Stein’s method, and Malliavin calculus underlie modern sharp results.

## 1. Classical and Entropic Berry–Esseen Theorems

The classical Berry–Esseen theorem establishes that for i.i.d. real random variables $X_i$ with $\E X_i=0$, $\Var(X_i) = \sigma^2 > 0$, and $\E|X_i|^3<\infty$, the normalized sum $W_n = S_n/(\sigma\sqrt n)$ satisfies
\[
\sup_{x\in\mathbb{R}} |\mathbb{P}(W_n \le x) - \Phi(x)| \le C \frac{\E|X_1|^3}{\sigma^3\sqrt{n}},
\]
where $\Phi$ is the standard normal CDF and $C$ is an absolute constant [2104.03479].

A major refinement is the *entropic Berry–Esseen bound* [1105.4119]. Suppose $\{X_k\}$ are independent, centered, with variances $\sigma_k^2>0$, and denote $B_n^2=\sum_{k=1}^{n}\Var(X_k)$. The total variation distance between the law $\mathcal{L}(S_n/B_n)$ and $\Phi$ admits
\[
\|F_n - \Phi\|_{\mathrm{TV}} \le C(D) L_3,
\quad L_3 = \sum_{k=1}^n \frac{\E|X_k|^3}{B_n^3},
\]
where $D(X_k) = h(N(0,\sigma_k^2)) - h(X_k)$ (relative entropy to normality) is uniformly bounded. Relative entropy bounds (Kullback–Leibler divergence) can be derived assuming fourth moments and uniform entropy control, with $O(n^{-1})$ rates sharper than Kolmogorov-type bounds.

Table: Comparison of Classical and Entropic Berry–Esseen Rates

| Setting           | Metric              | Moment Condition      | Rate (i.i.d.)    |
|-------------------|---------------------|----------------------|------------------|
| Classical         | Kolmogorov          | $\E|X|^3<\infty$     | $O(n^{-1/2})$    |
| Entropic          | Total Variation     | $\E|X|^3$, bounded $D$ | $O(n^{-1/2})$ | 
| Entropic          | Relative Entropy    | $\E[X^4]$, bounded $D$ | $O(n^{-1})$   |

The entropic approach replaces density boundedness with uniform entropy control and facilitates applications to weighted sums, $W_2$ bounds via transportation inequalities, and explicitly exhibits limitations for discrete laws.

## 2. Extensions: Dependence, U-Statistics, and Graph Functionals

Berry–Esseen analysis has been extended to sums with local or graph-based dependencies [2212.02590, 2104.03479, 2602.02217, 2104.03683]. For variables with a dependency graph $G=(V,E)$ of degree $\Delta$,
\[
d_K\left(W, \mathcal{N}(0,1)\right)\leq C\left((\Delta+1)^2 \sum_{k \in V} \frac{\E|X_k|^3}{v^3} + \cdots\right)
\]
highlights explicit dependence on both the graph structure and higher moments [2212.02590].

For generalized $U$-statistics and Hoeffding decompositions, optimal $O(n^{-1/2})$ (or $O(n^{-1})$ for degenerate/edge-only kernels) rates are attainable under mild fourth-moment conditions and connectedness of principal projections [2104.03479]. These results leverage innovative uses of exchangeable pairs under Stein's method, accommodating complex combinatorial dependencies (e.g., subgraph counts in random graphs).

In combinatorial/probabilistic settings, e.g. Rademacher functionals or permutational statistics, modern approaches (discrete Malliavin–Stein [1503.01029], stratified permutation CLT [2503.13986]) yield Berry–Esseen rates in Kolmogorov distance that are sharp up to constants and subtle combinatorial multiplicities.

## 3. High-Dimensional, Multivariate, and Self-Normalized Bounds

In high-dimensional regimes, Berry–Esseen theory quantifies both the rate and the explicit dependence on ambient dimension. For sums of independent/weakly dependent $d$-vectors $W=\sum X_i$ standardized such that $\E WW^T = I_d$,
\[
\sup_{\text{convex } A} |\mathbb{P}(W\in A) - \mathbb{P}(Z\in A)|\le C d^{1/4} \sum_i \E \|X_i\|^3
\]
with $d^{1/4}$ optimal in the purely linear case [2102.04923].
For general nonlinear statistics, $d^{1/2}$ dependence is currently the best known, and further refinements remain a research frontier [2102.04923]. For high-dimensional martingale difference sequences, recent work establishes $d_K$ rates over rectangles of $O(n^{-1/4})$ with only polylogarithmic dimension dependence, reflecting inherent limitations due to the third-moment regime [2605.03100].

Self-normalized Berry–Esseen bounds—the basis for inference using Studentized or variance-ratio statistics—are sharp: for martingale differences $\{X_i\}$,
\[
\sup_x \left| P\left(\frac{S_n}{\sqrt{[S]_n}} < x\right) - \Phi(x) \right| \le C_p N_n^{1/(2p+1)}
\]
where $N_n$ tracks $2p$-moments and the difference between sample and predictable quadratic variation, with $1/(2p+1)$ order shown to be optimal [1712.03667, 2104.03683]. Leung–Shao [2303.08619] provide nonuniform bounds for Studentized $U$-statistics, the first with valid nonuniform $1/\sqrt{n}$ decay modulo exponentially small correction, essential for bootstrap and tail probability control.

## 4. Local Limit Theorems and Density-Level Berry–Esseen Bounds

For precise proximity at the density level (“local” Berry–Esseen), optimal non-asymptotic results require even sharper control. In the multivariate case, under bounded densities $M$, variance $\sigma_k^2$,
\[
\sup_{x\in\R^d}|p_n(x)-\varphi(x)| \le (C\sigma)^d M^2 \lambda_3
\]
where $\lambda_3$ is the uniform directional Lyapunov third-moment, and extra $M^2$ reflects the smoothing cost of density-level control [2407.20744]. If all summands are symmetric, the fourth moment $B_4$ replaces $\lambda_3$, but at the cost of an extra $M$.

Log-concave, isotropic cases attain $O(n^{-1/2})$ for densities, but in general, the rate for densities (as opposed to distribution functions) is dominated by maximal density and Lyapunov coefficients.

## 5. Berry–Esseen Bounds Beyond the Classical Setting: Operator-Valued, Complex, and Non-Gaussian Limits

The extension of Berry–Esseen-type estimates into operator-valued free probability, complex Wiener–Itô integrals, and non-Gaussian limiting laws is now established.

For operator-valued free CLT,
\[
L(\mu_{X_n},\mu_{S_n}) \le c [A_1(x)]^{1/7} n^{-1/14}
\]
with $\|G_{X_n}(b)-G_{S_n}(b)\| \le n^{-1/2}$ at the resolvent/Cauchy transform level, where $A_1(x)$ aggregates operator-valued moment information [2105.02044].

In the complex multiple Wiener–Itô setting, optimal Wasserstein bounds in terms of kernel contractions and fourth-moment cumulants match those in the real case, with necessary and sufficient “fourth-moment contraction” conditions identified for the complex CLT [2407.05353]. Rates $T^{-1/2}$ replace classical $n^{-1/2}$ (as appropriate for Ornstein–Uhlenbeck processes in continuous time).

For isotonic regression and other non-regular settings, nonnormal Chernoff-type limits arise; Han–Kato derive Berry–Esseen rates matching those for oracle local average estimators, up to poly-logarithmic factors, via new anti-concentration inequalities and localization [1910.09662].

## 6. Methodological Innovations: Stein’s Method, Concentration, Malliavin Calculus

Stein’s method, including the exchangeable pairs and zero-bias couplings, underlies most modern Berry–Esseen advances. For dependent structures, innovations include stratified permutations [2503.13986], concentration-of-measure via randomized windows [2602.02217], and stochastic calculus on discrete/Malliavin spaces [2010.04387, 1503.01029]. The same tools are fundamental to both the classical bounds and extensions, with moment and contraction terms encoding the relevant combinatorics and dependence.

Non-asymptotic control—even in complex dependency structures, functional statistics (e.g., entropy estimators [2604.11489]), or operator-algebraic regimes—is now available, with explicit Berry–Esseen constants and rates in most contemporary inferential settings.

## 7. Summary Table: Berry–Esseen Bounds in Diverse Settings

| Structure                 | Metric          | Rate (sharpest)  | Key Condition(s)        | Reference         |
|---------------------------|-----------------|------------------|-------------------------|-------------------|
| I.I.D., sum               | Kolmogorov      | $O(n^{-1/2})$    | $\E|X|^3<\infty$        | [2104.03479]      |
| Non-i.i.d., entropy controlled | TV/entropy | $O(n^{-1/2})$, $O(n^{-1})$ | $\E|X|^3$/$\E X^4$, $D(X)<\infty$ | [1105.4119] |
| Sparse dependency graph   | Kolmogorov      | $O((\Delta+1)^2 n^{-1/2})$ | bounded third ($\delta\ge3$)  | [2212.02590]      |
| $U$-statistics            | Kolmogorov      | $O(n^{-1/2})$    | $\Vert f \Vert_4$<∞, connectedness | [2104.03479] |
| High-d martingale diff.   | Kolmogorov (rect) | $O(n^{-1/4} (\log d)^2)$ | thrid moments, cond. var. | [2605.03100] |
| Self-normalized sums      | Kolmogorov      | $O(n^{-1/2})$    | $4$th moments, local dep. | [2104.03683], [2602.02217] |
| Permutation/statistical design | Kolmogorov | $O(n^{-1/2})$    | $3$rd moments            | [2503.13986]      |
| Complex/operator-valued   | Wasserstein/Levy| $O(n^{-1/2})$ (resolvent), $O(n^{-1/14})$ (prob.) | moments, kernel contractions | [2105.02044], [2407.05353] |
| Isotonic regression       | Non-Gaussian    | $\sim n^{-1/3}$ (Chernoff) | model smoothness         | [1910.09662]      |
| Plug-in entropy/diversity | Kolmogorov      | $O(n^{-c})$      | H\"older regularity, tail | [2604.11489]      |

All rates are up to constants; actual rates and dependencies on structure-specific combinatorial or analytic parameters are explicit in the cited references.

---

**References:**  
- [1105.4119] Bobkov–Chistyakov–Götze, "Berry–Esseen bounds in the entropic central limit theorem"  
- [2104.03479] Shao–Zhang, "Berry--Esseen bounds for generalized $U$ statistics"  
- [2212.02590] Janisch–Le Héricy, "Berry-Esseen-type estimates for random variables with a sparse dependency graph"  
- [2602.02217] Cai–Shao–Zhang, "Refined Berry-Esseen bounds under local dependence"  
- [2605.03100] Shao–Zhang, "Berry-Esseen bounds for multivariate martingale difference sequences in the Kolmogorov distance"  
- [2104.03683] Shao–Zhang, "Berry--Esseen bounds for self-normalized sums of local dependent random variables"  
- [1503.01029] Krokowski–Reichenbachs–Thäle, "Discrete Malliavin–Stein method: Berry-Esseen bounds for random graphs and percolation"  
- [2503.13986] "Stratified Permutational Berry--Esseen Bounds and Their Applications to Statistics"  
- [2407.20744] Bobkov–Götze, "Berry-Esseen bounds in local limit theorems"  
- [2010.04387] Privault–Serafin, "Berry-Esseen bounds for functionals of independent random variables"  
- [2407.05353] "Berry-Esséen bound for complex Wiener-Itô integral"  
- [2105.02044] "Berry-Esseen bounds for the multivariate $\mathcal{B}$-free CLT and operator-valued matrices"  
- [1910.09662] Han–Kato, "Berry-Esseen bounds for Chernoff-type non-standard asymptotics in isotonic regression"  
- [2604.11489] Yu–Miao, "Berry-Esseen bounds for estimators of entropy and diversity indices on countable alphabets"  
- [1712.03667] Fan–Shao, "Berry-Esseen bounds for self-normalized martingales"  
- [2303.08619] Leung–Shao, "Nonuniform Berry-Esseen bounds for Studentized U-statistics"  
- [1710.01554] Shao–Zhang, "Berry-Esseen Bounds of Normal and Non-normal Approximation for Unbounded Exchangeable Pairs"

For proofs, constants, and further structural insights see the cited arXiv preprints.

Source: https://www.emergentmind.com/topics/berry-esseen-bounds