Rulers, Religion, and Riches: Why the West Got Rich and the Middle East Did Not (Cambridge Studies in Economics, Choice, and Society)
Page 33
36 Quoted in Hunt and Murray (1999, p. 65).
37 See Hunt and Murray (1999, p. 64) and Kohn (1999).
38 See Einzig (1970, p. 67).
39 See Hunt and Murray (1999). The operations of the Florentine Covoni family, who between 1336 and 1340 registered 443 exchange transactions, exemplifies the use of bills of exchange as a financial instrument: 70 were trade-related and 373 were financial (Mueller 1997, p. 317–18). Bills of exchange evolved further in the late sixteenth and seventeenth centuries when they became negotiable and endorsable (the first use of endorsement occurred in the 1570s). As endorsable instruments, bills were similar to convertible money (Kohn 1999).
40 Merchants eventually adopted bills quoted in fictitious units of stable value in order to escape changes in exchange rates resulting from currency debasement and speculation, but their adoption of this measure instead of discounting suggests that currency exchange maintained its important role in the exchange transaction (Einzig 1970). Another way that bills simulated interest-bearing loans was through non-repayment by the payer. In this case, it was tacitly understood by all parties that a dishonored bill would be protested in court (for appearances) and returned to its place of issue, after which the taker was obligated to pay the deliverer back at the current rate of rechange, which acted as an interest payment (Einzig 1970).
41 If lenders could use differences in exchange rates to make an arbitrage-like profit, why did markets not eventually clear and exchange rates equalize? Raymond de Roover (1944) suggests that differences in exchange rates reflected a built-in interest payment, and hence such differences had to exist for an equilibrium to hold. If no differences in exchange rates existed, then there would not have been incentive for the capital-wealthy to lend. Meanwhile, some merchants were willing to pay a premium to have access to this capital. For instance, sellers of bills in London were often merchants who needed access to cash to pay for cloth, which they expected to sell in the Low Countries. One way of gaining access to this credit was by selling a bill in London and honored in Antwerp or Amsterdam. See de Roover (1944).
42 The sakk and ruq’a acted like checks, and merchants employed them primarily in short-distance trade for relatively small sums (Goitein 1967, pp. 240–1; Udovitch 1975).
43 See Lieber (1968) and Udovitch (1979). Though it is certain that the suftaja predates the European bill of exchange, there is considerable debate concerning the Middle Eastern origins of the European bill. Early twentieth-century scholars such as Usher (1914) believed Western bills to be of Italian origin, while later “Orientalist” scholars such as Schacht (1964) and Lieber (1968) believe that European bills owe a great deal to the Islamic world. Ashtor (1973) reconciles the two viewpoints, noting that while Europeans were aware of suftaja and even dealt in them, the difference in the economic setting in which they emerged, which (as emphasized in this chapter) permitted an exchange transaction to be included in the European but not the Islamic bill, suggests that the European bill was a fundamentally different and unique credit instrument.
44 In a study of early safatij, Eliahu Ashtor (1973, p. 562) notes that “studying the texts referring to the suftadjas drawn up in Iraq and Egypt at the time of the Abbasid caliphs, we note that the sums sent to another city or another country had to be collected in the same type of money in which the loan was made” (italics added). The lack of a currency exchange associated with the suftaja extends well beyond the Abbasid period and is a salient feature of transactions registered in the Geniza in the twelfth and thirteenth centuries. Also see Udovitch (1975, 1979).
45 See Lieber (1968, p. 233), Ashtor (1973, pp. 556–7), Ray (1997, p. 71), and Pamuk (2004b).
46 Quoted in Goitein (1967, p. 243). Similarly, a characteristic “blank” suftaja read: “Give ____ all that he may demand, obtain a receipt from him, and debit the sum to me” (see Mez 1937, p. 476).
47 See Goitein (1967, p. 243).
48 The Hanafi permitted safatij.
49 An alternative hypothesis for the absence of an exchange transaction associated with the suftaja is that there were fewer opportunities to trade currencies, perhaps stemming from less fragmentation in the Middle East relative to Europe, and thus less scope to use currency exchange. Historical evidence indicating that numerous types of currencies, such as different types of dinars and dirhams, were available in the Middle East contradicts such a theory, however. Ashtor (1973, p. 560) notes that a “rich variety of money, that is to say the ease with which foreign monies could be obtained in the big cities, was a typical phenomenon of the monetary life of the Muslim countries at the time of the Abbasid caliphs and at that of the Crusades, distinguishing them signally, in this respect, from the countries of Western Europe.” Moreover, the fact that differences in exchange rates in Europe were essential to bills being profitable does not mean that such differences could not have emerged in the Middle East, if they indeed did not exist. Once financiers used European bills of exchange as instruments of finance, differences in exchange rates emerged endogenously as interest payments. It thus follows that had Middle Eastern lenders been able to include an exchange transaction with the suftaja, differences in exchange rates in different Middle Eastern cities may have followed.
50 A lender could buy a suftaja in place A, have an agent turn in the suftaja in place B for the same currency, have the agent exchange the currency for a different currency in place B, buy another suftaja in place B with the new currency, turn in that suftaja in place A in the new currency, and finally exchange the currency in place A for the original currency.
51 Two schools of Sunni Islam (Maliki and Shafi’i) explicitly forbade safatij, though the Malikites permitted their use in cases of extreme danger to the traveling merchant. The Hanbali school permitted them as long as no fee was charged. They were disapproved of, though permitted, by the Hanafi school (Dien 1995). The Hanafites, however, insisted that the suftaja was only permissible when there was no agreement to pay elsewhere and where the sums paid and repaid were equal (Ashtor 1973).
52 The enforceability of fines for late repayment suggests another possible mechanism that lenders could have secured a profit via safatij. The lender and borrower could have had a tacit agreement that the agent in the distant land would be late in repayment with the fee paid serving as interest. Indeed, Western Europeans employed this type of agreement. It is unlikely that Muslim lenders used this tactic for a variety of reasons, all of which are consistent with the theory presented in this chapter. First and foremost, this would have been a clear violation of Islamic law. While Muslim lenders used numerous hiyal that were consistent with the letter but not the spirit of the law, any implicit understanding between parties would have made the contract voidable under Islamic law. Thus, the essential difference between the two regions is that a dishonored bill would have been enforceable in Western European courts regardless of the intent of the parties. Indeed, the Church considered such an arrangement usurious (in fraudem usurarum, see Munro [2003]) but had little power to impose secular sanctions after its power waned in the late thirteenth century, whereas such a bill would not have been enforced in Islamic courts if it were obvious that the intent was to circumvent interest restrictions. Moreover, even if such a practice became widespread, it is unclear how it would have facilitated impersonal lending. The set of potential sanctions that could enforce this type of contract were personal or social.
53 See Einzig (1970).
54 See de Roover (1946b, 1963). The Medici house operated in a similar manner to its rival controlled by Francesco Datini. The Medici enterprise differed from the “super-company” organizations of the fourteenth century (such as the Peruzzi, Bardi, and Acciaiuoli companies), which were centralized under one partnership that controlled foreign branches.
55 See de Roover (1963, p. 87).
56 See de Roover (1946a, 1963).
57 See Goitein (1967, pp. 244–5) and Udovitch (1975).
58 Theoretically, Middle Eastern lenders could have extended th
eir networks in order to increase their confidence in the partner on the other end of the transaction, who would have been a part of the same “business.” This would have encouraged the writing of larger safatij at greater fees. Yet, in this case the incidence of personal exchange is even greater, as both the borrower and his agent are part of an even closer network. It is also possible that Middle Eastern lenders could have learned the potential benefits of adding exchange to the suftaja through contact with Christian minorities. Indeed, the Pact of Umar permitted Christian minorities (dhimmis) in Islamic lands to utilize Christian courts in transactions involving non-Muslims. Yet, it is unlikely that European bills of exchange could have been commonly employed as financial instruments in Muslim lands for two reasons: (1) bills of exchange were enforceable only by merchant law in Europe, which was not available in Islamic law; and (2) the viability of bills of exchange as financial instruments depended on the existence of a critical mass of (in this case, Christian) borrowers and lenders in more than one region. In fact, Christian minorities generally abided by Islamic law until the eighteenth century, by which time much more advanced financial instruments were available to European lenders. For more on these points, see Kuran (2004a, 2011).
59 See Mokyr (1990). For more on the importance of historical events and path dependence on the evolution of institutions, see David (1994), Kuran (2005a, 2011), and Greif (2006b, chs. 5, 7).
5 Restrictions on the Printing Press
1 I am only concerned here with the invention of the movable-type printing press in Europe. The Chinese knew of printing since the eleventh century, but it was not introduced to Europe until the 1450s.
2 See McCusker (2005) and Chilosi and Volckart (2010).
3 The actual number spans the period 500–1450, but it is almost certain that the number of manuscripts produced from 450 to 1450 was smaller than in the half-century following the invention of the press.
4 See Febvre and Martin (1958, p. 218).
5 See Spitz (1985) and Buringh and van Zanden (2009).
6 See Buringh and van Zanden (2009).
7 See Febvre and Martin (1958).
8 See Dittmar (2011).
9 See Eisenstein (1979).
10 See Febvre and Martin (1958, p. 249).
11 See Febvre and Martin (1958) and Eisenstein (1979).
12 See Swetz (1987).
13 Quoted in Swetz (1987, p. 25).
14 See Kertcher and Margalit (2006).
15 For more, see Febvre and Martin (1958), Eisenstein (1979), Love (1993), Johns (1998), and Kertcher and Margalit (2006).
16 Much of the next two sections are in Coşgel, Miceli, and Rubin (2012a). I thank Metin and Tom for their work and their permission to let me use the ideas we formulated together in this chapter. And I especially thank Metin for letting me use the Turkish works he translated.
17 Mystakidis (1911, p. 324) mentioned the presence of such an edict in the first volume of Türk Tarih Encümeni Dergisi, but the validity of this claim was challenged by Efdaleddin Tekiner (1916) in the same publication five years later on the grounds that Ottoman archives do not house edicts issued prior to 1553 and thus Mystakidis could not possibly have seen it. Despite this correction and the fact that no such edicts have since been uncovered, the secondary literature has for the most part accepted the presence of the edict as a matter of established fact. See, for example, Pedersen (1984, p. 133), Finkel (2005, p. 366) and Savage-Smith (2003, p. 656). I thank Metin Coşgel for the insights and translations on Mystakidis and Tekiner.
18 English translation from Göçek (1987, p. 112).
19 See Finkel (2005, p. 366).
20 For more see Coşgel, Miceli, and Rubin (2012a) and Frazee (1983).
21 See Pedersen (1984, p. 135).
22 See Finkel (2005, p. 366).
23 See Atiyeh (1995, p. 285).
24 See Göçek (1987, p. 110).
25 For more, see Pedersen (1984), Robinson (1993, p. 233), Sardar (1993), and Atiyeh (1995, p. 283).
26 On the Ottoman Empire, see Quataert (2000, p. 167). On Europe, see Baten and van Zanden (2008).
27 See Özmucur and Pamuk (2002).
28 See Sardar (1993, pp. 47–51).
29 See Sardar (1993, pp. 47–51).
30 See İnalcık (1973, p. 99) and Dale (2010).
31 See Hourani (1991, ch. 13), Imber (1997), and Dale (2010).
32 This process is described in Hourani (1991, p. 199) and Robinson (1993, p. 235).
33 See Sardar (1993, p. 50).
34 See Sardar (1993, pp. 45–6).
35 See Robinson (1993, p. 237).
36 Quoted in Robinson (1993, p. 237).
37 See Sardar (1993, pp. 52–3).
38 Chaney (2016) puts forth a similar argument. He argues that the fall of Muslim science was due to the fact that Islamic religious authorities faced little competition in the realm of ideas and hoped to keep it that way.
39 See Göçek (1987, p. 109).
40 See Buringh and van Zanden (2009).
41 See Tierney (1988).
42 See Schachner (1962).
43 See Schachner (1962, p. 50).
44 See Christ et al. (1984, pp. 297–310) and Febvre and Martin (1958, pp. 22–5).
45 See Haskins (1957, pp. 38–53), Schachner (1962), and Christ et al. (1984, pp. 237–8).
46 To be clear, the Ottoman suppression of printing in the Arabic script was not the result of idiosyncratic decisions made by a few sultans in the fifteenth and sixteenth centuries. If this were the case, it is likely that the long hand of history would have caught up with the Ottomans and pushed toward the adoption of printing. Instead, as this chapter suggests, the Ottoman blocking of printing was a calculated decision resulting from very deeply entrenched, institutionally imposed incentives faced by the sultan. The Ottoman non-adoption of the press was a self-enforcing equilibrium outcome, not a choice made by a few foolish sultans.
6 Printing and the Reformation
1 Spenkuch (2016) provides possible support for Weber’s hypothesis. He finds that Protestantism induces people to work longer hours – leading to higher earnings – and that human capital or institutional differences cannot account for these results. This leaves many possible causal channels open, including the one proposed by Weber.
2 Becker and Wößmann (2008, 2009) propose that the causal pathway connecting Protestantism to long-run economic success is education. They argue that Luther encouraged reading the Bible, which in turn gave Protestants an early start on acquiring literacy. Arruñada (2010) argues that Protestants did not have a unique work ethic, but instead had a “social ethic” that favored market transactions. Young (2009) overviews a number of possible, non-mutually exclusive reasons that Protestant regions had better long-run outcomes than Catholic regions. Guiso et al. (2003) gives a contrary view. They find a positive correlation between Christian religions and attitudes conducive to economic growth.
3 As calculated in Allen (2001).
4 Allen (2001) gives data for 1900–1913, but Bairoch et al. (1988) does not provide population data for this period. Allen (2001) also does not have data for each city in each period. Where Allen’s data are missing, I exclude these cities from the analysis. Population data from Bairoch et al. (1988) are from the beginning of the period in question. Bairoch and colleagues do not report population data for 1550 or 1650, so I derive these data by taking the geometric mean of the two surrounding points.
5 On the Reformation, see Weber (1905 [2002]), Tawney (1926), Becker and Wößmann (2008, 2009), and Arruñada (2010). On the printing press, see Eisenstein (1979), Baten and Van Zanden (2008), Buringh and van Zanden (2009), and Dittmar (2011). On the New World, see Pomeranz (2000) and Acemoglu, Johnson, and Robinson (2005). On the Renaissance, see Mokyr (2002) and McCloskey (2010). On the Ottomans, see Iyigun (2008, 2015).
6 Rubin (2014a) suggests that centralized institutions like the medieval Church are particularly vulnerable to rapid revolt because they have numerous means of suppr
essing dissent. This means that the publicly stated preferences of people often differ from their privately held preferences (as in Kuran [1995]). Makowsky and Rubin (2013) further this argument, suggesting that information technology further increases the likelihood of revolt in economies with centralized institutions, as previously suppressed anti-authority preferences are more likely to rise to the surface.
7 Much of this section is from Rubin (2014b).