Response to Carlos Marichal and William Summerhill

We thank Carlos Marichal, William Summerhill, and the editors of the HAHR for taking up the debate on how to assess the governance structures under lying Spanish rule in the Iberian Peninsula and the American colonies and (more generally) the current use of institutional explanations in social sciences and history as determinants of differential economic development. Both commentators seem perfectly comfortable with our actual numbers and our attempt to draw attention to the understudied intracolonial transfers, so we will waste no time restating our basic findings. Yet, with great insight Marichal and Summerhill question several important aspects of our interpretation of these features. The issues raised are large, but in the interest of being concise we will offer only a few short comments for future debate.Although Marichal agrees with our critique of the characterization of Spanish absolutism as all-powerful and rigid, he in turn feels that our concepts of negotiation and consensus are too rigid. We concur that these require careful definition. In our view, negotiation and consensus in the Spanish case more often than not took the form of collusion and co-optation or corruption, as described by other historians.1 What is abundantly clear to us is that in the Spanish Empire, the nexus between ruler and subjects — individuals or corporate bodies — was not a relation of command.Still, Marichal argues that especially in the late colonial period, “coercion and censorship [were] powerful and frequently used instruments of the crown and the powerful to maintain the status quo.” To begin with, it is interesting that Marichal here automatically subsumes into one category “the crown” and “the powerful,” presumably colonial elites.2 For the crown or the elites, it was not easy to establish a systematic recourse to repression without an equivalent investment in some means of enforcement, a price the crown was not willing or able to pay, as John Coatsworth has recently pointed out and as we can confirm with a glance at the low spending on colonial bureaucracies and standing armies evident in the fiscal accounts.3 Thus any repression at all required the co-optation of and more often the initiative from local elites.4Marichal cites Scarlett O’Phelan Godoy to argue that the repeated episodes of protest and open revolt witnessed in eighteenth-century Spanish colonies were tax revolts caused by resentment against the mita, alcabalas, tariffs, and repartimiento. This is certainly true. But it is a long way from there to Marichal’s (implied) assertion that the rebels actually questioned Spanish rule per se rather than local power and abuse. Our reading of the recent literature suggests very little evidence for that. As Sergio Serulnikov has shown in his masterly study on the political culture of Andean peasantry, the rebellions were subverting the colonial order, not questioning its existence.5 Part of the motivation for our paper was to raise awareness of the need to distinguish clearly between colonial regime, order, and rule and between power elites, crown officials, peninsular councils, local institutions (Spanish, creole, or indigenous), and the crown.6 Eighteenth-century rebellions were often ignited by changes in the established mode of tax collection or the new assessment of existing contributions, which affected the tax burden of different groups, but the revolts did not seek a change of the rule.7 For example, the 1765 urban rebellion in Quito “showed few signs of seeking fundamental changes in the political and social order.”8 Equally, the goal of the Comuneros’ revolt of 1781 was to “persuade the king to repudiate the policies of his ministers.” As John Leddy Phelan put it, “The world without a monarchy was inconceivable in 1781.”9Marichal also takes issue with our characterization of the colonial fiscal system as “autonomous but interdependent fiscal districts . . . organized into a rather loose network.” His main counterargument is the traditional explanation for the existence of intracolonial transfers, namely that they were simply military subsidies from tax-rich to tax-poor parts of the colonies. On the one hand, we have shown in a recent article that the military motive explains less than half of the actual intracolonial transfers.10 On the other hand, that expenditures were designated to military ends in the books does not reveal the means through which these monies were allocated. With a slim bureaucracy and minimal expense for enforcement, how could the empire direct, execute, and control the spending other than by outsourcing it to merchants?11 Scholars of imperial defense spending have repeatedly shown how the incipient military presence fostered the early mercantile development of the cajas that received the subsidies.12 Spending resulted in a chronic disequilibrium vis-à-vis the (sending) royal treasuries. Over time, loans, cash advances, goods, and private bills appeared as instruments of payment to defray military expenditure, which could rarely be repaid in full by the situados. By the late eighteenth century the Spanish defense apparatus was in the hands of local merchants and manned by local militias.13 Military necessity alone can explain neither the sheer size of transfers in the late eighteenth century nor why elites in the paying regions acceded to the disbursement of money to be spent elsewhere.In a new paper we have explored the part of the story that is missing at present, namely the expenditure side, military or otherwise.14 As Marichal suggests, early modern states devoted about 80 – 90 percent of their expenditure to the army, navy, and war debts. In the late eighteenth century, during peacetime, Britain for example spent 31 percent of its total expenditure on the military (including wages) and 56 percent on the interest of largely war-related public debt.15 In our 1785 – 89 sample of Spanish American cajas, the percentage of comparable direct military expenditure is similar: 28 percent. But the fundamental difference is that the cajas owed little interest on old war debt.16 The point is not that the colonial cajas made no debts, but that these debts were incorporated into contracts with colonial merchant interests, which thus benefited greatly from the fiscal setup.17 Not surprisingly, the very fact that a caja was located in a large port town with powerful merchant elites — whose own businesses often transcended the boundaries of the district of their residence — increased its income from transfers considerably, independently of any military expenditure.18The crucial question remains how to explain the massive intracolonial transfers. Marichal suggests that the metropolitan government was able to force colonial elites to surrender large amounts of revenues to be spent outside their districts for military purposes. Using the fiscal accounts, we have shown that military expenditure alone cannot explain the transfers and that the share of expenditure spent on enforcement and repression was simply too small to support the assumption that the crown could have executed such massive redistribution single-handedly, even if it had wanted to. A colonial state that was famously ineffective in its policy implementation was unlikely to be able to enforce the one thing that all classes anywhere tend to resist, that is, being taxed for expenditures over which they had no control and from which they derived no benefit. The name of the game was fundamentally co-optation and consensus, not coercion and censorship, which could only be used sparingly.Like Carlos Marichal, William Summerhill is willing to accompany us part of the way, albeit on a different section of our path. He seems to be in agreement with our attempt to redefine what absolutism meant in practice, in particular that absolutism in the Spanish case was in many spheres antagonistic to the notion of centralization. In Freedom and Growth the late Larry Epstein showed clearly how “jurisdictional fragmentation” was part and parcel of early modern monarchical regimes.19 But ultimately Summerhill draws a line in the sand when it comes to our more general argument that limitation through an English-style parliament might matter less than generally assumed precisely because other formal and informal institutions might imply similarly strong constitutional constraints.Summerhill’s criticism reflects the ongoing debates about the link between political regimes and economic growth originating in North and Weingast’s influential 1989 article about the supposed fiscal impact of the English Glorious Revolution.20 The argument advanced was simple and attractive: monarchs who were not constrained by a parliament with a tax and spending veto suffered a credibility problem with potential moneylenders. Since monarchs could choose at will not to repay or could expropriate their lenders (and subjects) directly or through inflation, lenders would want to be compensated for the higher risk through a higher interest rate. As proof, North and Weingast cited the substantial decline in interest premiums on loans taken out by the English government after the Glorious Revolution. Though the total amount of debt increased dramatically, investors apparently began to have more trust in the government’s commitment to honor its debt.Summerhill refers at several points to the literature that has developed in the wake of this argument. We are told categorically that absolutists were “prone to impressments, confiscations, and default; [hence] absolutists confronted low debt ceilings, credit rationing, and had to pay hefty risk premiums on loans.” Summerhill rightly points out that we should think about fiscal discretion as a continuous variable and analyze where on such an index different rulers were located.21 But he adds, “At one extreme would be absolute monarchs . . . , who enjoyed great discretion over expenditures. At the other . . . would be strongly limited governments, where the crown’s influence over spending was completely subordinated to a standing legislative body.” (Interestingly, to the best of our knowledge no English parliament to date has ever stopped a government from starting a war, the largest possible outlay of money. And once a war was under way, refusing to allow spending was politically difficult.)Summerhill admits that North and Weingast’s original conclusions have been controversial. Epstein showed that the English state before the Glorious Revolution paid much higher interest rates that anyone else (republican or monarchic) in Europe. He pointed out that the Dutch fiscal know-how adopted hereafter in England might better explain investors’ greater confidence in the English fiscal system than the new powers of Parliament. The English fiscal machinery was simply catching up with its neighbors.22 To be sure, few states were as prone to default on their debt as was Spain. But this always presented historians with a paradox that directly challenged the North and Weingast proposition. The principal of one of the main consolidated debt instruments, the juros, increased from about 2.3 million ducados in 1504 to 80 million in 1598, and the crown defaulted three times in this period. But the effective interest rate charged on the juros fell from 10 percent to 5.8 percent over the same period, which in the North and Weingast model would indicate that confidence in the crown was rising.23North and Weingast’s idea that interest rates paid by the crown reflected the credibility of the ruler to honor sovereign debt fails precisely to account for jurisdictional fragmentation. Much of Spanish government debt in the metropolis and the Indies (like that of the Netherlands) was secured directly against specific taxes raised in determined regions or cities on individual products or particular taxpayers. The lender cared about the health of the wool trade on the Cantabrian coast rather than the monarch’s financial dire straits. Equally, the continued willingness of colonial merchants to advance goods and money (and therefore credit) to the authorities depended on the expected revenues from alcabalas or the tobacco monopoly, not on the solvency of the Real Hacienda. And with extended tax farming and privatized public expenditure, often the same person was tax collector, provider, and lender.This leads us back to our basic Epsteinian argument. Fiscal fragmentation in itself implied a whole set of “constitutional” constraints on the Spanish crown. Some of these were as formalized as any potential veto by an English parliament. Revoking the Basque fueros would have required a constitutional change that no king could undertake unilaterally. Other constraints were part of the unwritten contract between crown and colonial and metropolitan subjects.24 And yet English historians would rightly be adamant that an unwritten constitution might be just as effective as a written one. “Formal rules” and “informal discretion” are historically not as clearly distinguishable as Summer-hill suggests, but they are equally capable of limiting the ruler.North and Weingast’s framework has failed the historical test for the early modern period for at least two reasons. First, their definition of limited government as equal to English-style parliaments misses the point that existing constitutional constraints elsewhere also resulted in limited government. Hence, their conclusion that the Glorious Revolution caused the English financial revolution and superior economic performance is spurious because the basic assumptions of their model are flawed. Second, if any single factor could account for England’s exceptional performance, it was its centralization. England achieved its fiscal unification in 1066, seven hundred years before that of any other European state. This was Epstein’s conclusion from the Italian experience, and Dincecco’s recent cross-country sample confirms it for the pre-1815 period.25 Notably, Spain was still struggling with fiscal unification in the mid-nineteenth century and it was the country where Napoleonic centralization was arguably least successful.We thus agree with Summerhill that more formal modeling of the way in which early modern rule worked is necessary. But we have to draft the basic assumptions of such models with care. Most of the game-theoretical literature starts precisely from the North and Weingast point that rulers behave as predators who will try to maximize their revenue unless they are kept in check by parliament.26 However, in our view the Spanish path of nation-state and empire building was of a different nature. Spanish rulers maximized something more sophisticated than revenue, namely the survival of their rule and the aggrandizement of the empire. That involved raising taxes, to be sure, but always within the limits of what could be done without risking the subjects’ revocation of the contract.27 That was the lesson learned in the Netherlands and in Catalonia, and it was applied throughout the empire. And this explains why this empire was very successful in keeping its rule unchallenged with very low investment.As Summerhill intimates, intracolonial transfers might not have been an arrangement that maximized economic growth in the empire, and to the best of our knowledge we did not claim that. We deliberately refrained from making any judgements about how the governance structures that we characterize impacted aggregate economic activity. But we observe that regions which were the largest beneficiaries of intracolonial transfers in the late eighteenth century (Buenos Aires, Havana, Cartagena) are identified by other economic historians as the ones that grew fastest at this time.28What we do instead is to reverse the causality established between independence and the ensuing poor economic development of the emerging Latin American republics. When taking at face value the political-economic implications of the supposedly “absolutist” Spanish rule, historians are in danger of misinterpreting the cause and aftermath of independence. Few historians wonder why South America kept on fighting — as Spain did — for a good part of the nineteenth century, even well after independence had been achieved, or what regional elites within these republics fought over for decades. We advance a hypothesis that neither Summerhill nor Marichal seems to dispute. If our story is true, then the institutional explanation that anchors the miserable economic trajectory of Latin America in postindependence political disorder and instability needs a serious rethinking.Brazil’s independence from Portugal is probably the best counterfactual to our interpretation of fiscal and economic causes for the political instability in Spain and Spanish America after 1808. While the Spanish Empire imploded and fragmented, the Portuguese captaincies remained united both during and after Dom João’s 14-year stay in the New World colony.29 Brazil maintained the fundamental political, military, and economic institutional setup of the captaincies after separating from the metropolis in 1822. Only in the late 1880s did the political constitution of Brazil suffer a definitive transformation, at the same time at which slavery was abolished.30What was different in the case of Spanish American independence is the exogenous shock that crumbled the foundations of the Spanish state and empire.31 Napoleon’s imprisonment of the Spanish king undermined the very core of the Spanish rule, namely the established system of governance, the source of legitimacy of contracts, and the role of the king as ultimate arbiter in the system of intra-imperial subsidies. With a ruler deprived of his ability to rule, the need to endow legitimacy to the subsequent rules of the game (that is, the state) turned the ruler’s prerogative (the exercise of state power) into the arena for the protracted fiscal and political strife that prevailed throughout the nineteenth century in the former metropolis and colonies

Environmental ethics · Culture, Economy, and Development Studies · Historical Economic and Social Studies · Philosophy · Politics and Society in Latin America

• A stakeholder empire

  Open Access•Regina Grafe, Alejandra Irigoin•The Economic History Review•2012

  Add to list