关于「实证闭环:多店营业额合规自查、门槛聚束估计器与预注册检验」,《门槛聚束与营业额弹性》给出的定量结论是什么?

结论:第八章给出多店营业额合规自查流程、蒸馏小工具「门槛聚束估计器」与预注册检验方案:一旦发现异常堆积,交由会计师核对真实申报,而不是调整申报。书中含 22 段可运行的 Python 程序。

数据口径:书中参数为示意,门槛一律记作 T;全部计算由书中程序生成。 第一章在一份固定种子的示意分布上估出超额质量 942.0 家、观察弹性 0.1378,结构弹性为 0.25,差距来自摩擦与反事实近似。

English version

对照数据表

下表数字由本章 Python 程序原样打印,可重新运行复现。 程序只用 Python 标准库

level per store (Bonferroni): 0.00417
storebelowabovep0p-valueaction
S01010.7491.00000ok
S02240.6730.98371ok
S03210.7320.82336ok
S04620.5830.28155ok
S05680.5640.90157ok
S061370.5530.25992ok
S075100.4980.93875ok
S08960.4830.25846ok
S092000.4750.00000REVIEW
S10250.4390.88696ok
S11120.3850.76729ok
S12320.4000.31679ok
REVIEW means: reconcile declarations with till and bank
records through an accountant; never edit figures to pass
查看完整验算程序
# Chapter 8. p81: a multi-store compliance self-check.
# 12 stores, 60 periods each, declared turnover per period; T = 100
# units is an illustrative threshold. Two stores (S04, S09) contain
# declarations that were pushed just below T. For each store, a
# lognormal is fitted from the median and the median absolute
# deviation of log turnover; it gives the expected share p0 of the
# band [0.95T, 1.05T) that should fall below T. An exact binomial tail
# with a Bonferroni level flags stores for an accountant's review.
import math
import random

SEED, STORES, PERIODS, T = 20261113, 12, 60, 100.0
PUSHED = (4, 9)


def ncdf(x):
    return 0.5 * (1 + math.erf(x / math.sqrt(2)))


def median(xs):
    s = sorted(xs)
    m = len(s) // 2
    return s[m] if len(s) % 2 else (s[m - 1] + s[m]) / 2


def tail(k, n, p):
    return math.fsum(math.comb(n, j) * p ** j * (1 - p) ** (n - j)
                     for j in range(k, n + 1))


def store(rng, i):
    mu = 70.0 + 60.0 * (i - 1) / (STORES - 1)
    out = []
    for _ in range(PERIODS):
        z = mu * math.exp(rng.gauss(0.0, 0.15))
        if i in PUSHED and T < z <= 1.12 * T and rng.random() < 0.8:
            z = T * (1 - rng.uniform(0.0, 0.03))
        out.append(z)
    return out


if __name__ == "__main__":
    rng = random.Random(20261113)
    alpha = 0.05 / STORES
    print("level per store (Bonferroni): {:.5f}".format(alpha))
    print("store  below  above   p0     p-value   action")
    for i in range(1, STORES + 1):
        zs = store(rng, i)
        lg = [math.log(z) for z in zs]
        m = median(lg)
        s = 1.4826 * median([abs(x - m) for x in lg])
        cdf = [ncdf((math.log(f * T) - m) / s) for f in (0.95, 1.0, 1.05)]
        p0 = (cdf[1] - cdf[0]) / (cdf[2] - cdf[0])
        lo = len([z for z in zs if 0.95 * T <= z < T])
        hi = len([z for z in zs if T <= z < 1.05 * T])
        p = tail(lo, lo + hi, p0) if lo + hi else 1.0
        print("S{:02d} {:7d} {:6d} {:6.3f} {:10.5f}   {}".format(
            i, lo, hi, p0, p, "REVIEW" if p < alpha else "ok"))
    print("REVIEW means: reconcile declarations with till and bank")
    print("records through an accountant; never edit figures to pass")

唯一数据源与引用

范雅福(Fan, Y.). (2026). 《起征点聚束估计与营业额分布、门槛下方堆积揭示的行为弹性母典》第 8 章「实证闭环:多店营业额合规自查、门槛聚束估计器与预注册检验」(版本 1.0.0). 5D Library. DOI 待发布

DOI 待发布 · 专著页正式发布后上线(本站暂无专著页) · 5DL-B-319 · v1.0.0 · CC-BY-NC-ND-4.0

BibTeX
@book{fan2026319,
  author    = {Fan, Yafu},
  title     = {Bunching at Tax Thresholds and the Distribution of Turnover: Behavioural Elasticities Revealed by Piling-Up below the Line},
  year      = {2026},
  publisher = {5D Library},
  edition   = {1.0.0},
  note      = {DOI pending}
}

本书其他章节结论