5D Library › 分拆经济学、防套利身份锁定与多阶授权阶梯 · English

第七章 算缗告缗、《管子》与宋代盐钞

为什么在这里读这些文献

本书的三个问题——按客群分级收费、防止私下转手套利、用可转让的凭证特许经营——在中国古代财政中都有先例。本章读三份文献:《史记·平准书》所记汉武帝时的算缗与告缗,《管子·乘马》论市场与价格的一句,以及《宋史·食货志》中的盐钞,后者以 Chien 的英译注本为据。引文均已与所列底本逐字核对。

《史记·平准书》:分级税率与「以其半畀之」

原典

《史记·平准书》(中国哲学书电子化计划所录本)

諸賈人末作貰貸賣買,居邑稽諸物,及商以取利者,雖無市籍,各以其物自占,率緡錢二千而一算。諸作有租及鑄,率緡錢四千一算。……匿不自占,占不悉,戍邊一歲,沒入緡錢。有能告者,以其半畀之。

楊可告緡遍天下,中家以上大抵皆遇告。杜周治之,獄少反者。乃分遣御史廷尉正監分曹往,即治郡國緡錢,得民財物以億計。

(Sima Qian, n.d.)

这段文字包含三个与本书同构的设计。其一是分级税率:商贾每二千钱一算,有租税的手工业与铸造者每四千钱一算,按经营性质把纳税人分成两档,与第五章按客群分级定价是同一种思路。其二是自行申报加惩罚:不申报或申报不全,罚戍边一年并没收缗钱。其三是告发者分一半:「有能告者,以其半畀之」,告发者得到没收财产的一半,这与第三章 50% 的分销佣金是同一个比例。

推导

# Chapter 7. [p14] The numbers of the classics, recomputed.
# (1) Shiji, Pingzhun shu: merchants pay one suan per 2,000 cash of
# declared wealth, craftsmen one per 4,000; concealment is punished by
# a year at the frontier and confiscation, and an informer receives
# half ("yi qi ban bi zhi"). How many cash one suan held is left as a
# parameter c. A merchant who hides wealth W is caught with probability
# q; hiding pays (ignoring the frontier year) when q W < W c / 2,000.
# (2) Songshi salt treatise (Chien 2004): old vouchers were bought back
# at 30 percent of face value. If a voucher's worth is the salt behind
# it divided by the claims issued, that is 1 / 0.30 claims per unit.
# (3) The same arithmetic for a card with k extra slots: k + 1 claims.

if __name__ == "__main__":
    print("   (1) break-even detection probability q* = c / base")
    print("       c (cash per suan)   merchants (2,000)  craftsmen (4,000)")
    for c in (60, 120, 240):
        print("       {:17d} {:18.3f} {:18.3f}".format(
            c, c / 2000, c / 4000))
    print("       informer's share of confiscated wealth: {:.2f}".format(
        1 / 2))
    print("   (2) buy-back at 0.30 of face -> {:.2f} claims per unit"
          .format(1 / 0.30))
    print("   (3) card claims per payment: " + "  ".join(
        "k={}: {}".format(k, k + 1) for k in range(0, 5)))
    print("\nReading: the Han rule made concealment a bet on a small")
    print("detection probability and paid informers half, the same")
    print("ratio as a 50% affiliate commission. If a voucher's worth were")
    print("the salt behind it over the claims issued, a buy-back at 30%")
    print("would mean more than three claims per unit of salt - as a card")
    print("with k = 2 extra slots issues three claims for one payment.")
#    (1) break-even detection probability q* = c / base
#        c (cash per suan)   merchants (2,000)  craftsmen (4,000)
#                       60              0.030              0.015
#                      120              0.060              0.030
#                      240              0.120              0.060
#        informer's share of confiscated wealth: 0.50
#    (2) buy-back at 0.30 of face -> 3.33 claims per unit
#    (3) card claims per payment: k=0: 1  k=1: 2  k=2: 3  k=3: 4  k=4: 5
#
# Reading: the Han rule made concealment a bet on a small
# detection probability and paid informers half, the same
# ratio as a 50% affiliate commission. If a voucher's worth were
# the salt behind it over the claims issued, a buy-back at 30%
# would mean more than three claims per unit of salt - as a card
# with k = 2 extra slots issues three claims for one payment.

汉代一算合多少钱,本书不作断定,程序把它当作参数 cc。商贾隐匿财产被告发的概率若低于 c/2000c/2000,隐匿就比申报划算(不计戍边之苦);cc 取 120 时这个门槛是 0.060,手工业者是 0.030。这样低的门槛意味着,单靠官府稽查很难让人如实申报,所以要把一半的没收所得分给告发者,让每一位邻人都成为稽查员。结果如《平准书》所记:「中家以上大抵皆遇告」,「得民財物以億計」。用本书的语言说,告缗令把一个激励不相容的申报制度,改成了一个人人有动机举报的制度,代价是社会信任的大面积破坏。第四章的模仿动态也提醒我们,高额分成会以同样的方式在渠道中扩散。

《管子·乘马》:市者,货之准也

原典

《管子·乘马》(Project Gutenberg 电子本)

地者,政之本也。朝者,義之理也。市者,貨之準也。黃金者,用之量也。

(Anonymous, n.d.)

「市者,货之准也」:市场是衡量货物价值的标准。第一、二章的全部计算都依赖这一点:最优价格只能从买家的估值分布中求出,而估值分布只能从市场上观察。一个卖家若不观察市场就给全通卡定价,正是在没有「准」的情况下定价。

宋代盐钞:凭证的超发与贬值

据 Chien 对《宋史·食货志》盐法部分的英译与注释,范祥(?—1060)任陕西提点刑狱兼制置解盐,于 1048 年改革解州池盐的管理,取消官卖,改由商人凭新的「钞引」运销;商人在榷货务以现钱、金银或纸币购钞,每张钞对应一定数量的盐,与盐价无关(Chien 2004,第 63—64 页)(Chien 2004)。此后朝廷屡次大量发钞,钞的价值随之下降(第 65 页);神宗时旧钞按原值的 30% 收回(第 66 页注);蔡京当政时,商人被迫一再以亏损换钞,往往要缴三次费用才拿到盐(第 69 页)。南宋初年,朝廷发行盐、茶、酒、醋等钞引筹钱,Chien 记载「Alert merchants pursued the retinue to snap up these vouchers」(导言)。

程序的第二、三项把这一经验与第三章联系起来:若一张凭证的价值等于它背后的盐除以发出的凭证数,按 0.30 收回意味着每单位盐对应了 3.33 张凭证;而一张附赠 k=2k = 2 个名额的个人年卡,就是用一次付款发出了三张「凭证」。

假说 · 尚未验证

作者的诠释(尚未验证): 附赠名额与宋代钞引超发是同一种机制:把对同一份产能的索取权发得多于产能,索取权就会在二级市场上折价流通,并把利润从发行者转给中间商。什么会证明这一诠释不当:若宋代盐钞贬值的主要原因被证明不是发行数量,而是盐的产量或运输条件的变化,那么这一类比只在形式上成立。

小结

《平准书》的算缗令按经营性质分级征收,并把没收财产的一半分给告发者,与本书的分级定价和 50% 佣金同构;它也显示了高额分成如何改变整个社会的行为。《管子》「市者,货之准也」提醒定价必须以市场估值为准。宋代盐钞的超发与贬值,是附赠名额被转售的历史版本。

参考文献

Anonymous. n.d. Guanzi, Chapter 5: Cheng Ma. In Chinese; Project Gutenberg e-Text No. 7367.
Chien, Cecilia Lee-fang. 2004. Salt and State: An Annotated Translation of the Songshi Salt Monopoly Treatise. Michigan Monographs in Chinese Studies 99. Center for Chinese Studies, The University of Michigan.
Sima Qian. n.d. Shiji [Records of the Grand Historian], Juan 30: Pingzhun Shu [Treatise on the Balanced Standard]. In Chinese; Text of the Chinese Text Project, Ctext.org/Shiji/Ping-Zhun-Shu.

Fan, Y. (2026). Unbundling Economics, Anti-Arbitrage Identity Locking and Multi-Tier Licensing Ladders: Orthogonal Demand Partitioning, Resale Cannibalization Bounds, Affiliate Margin Preservation and Five-Figure B2B Upsell Equilibria [分拆经济学、防套利身份锁定与多阶授权阶梯] (Version 1.0.0). 5D Library. https://doi.org/10.5281/zenodo.24090184 · Licensed under Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0).