Anatomy of a premodern state

Abstract

We provide a blueprint for constructing measures of state capacity in premodern states, offering several advantages over the current state of the art. We argue that assessing changing state capacity requires considering the composition of revenues, expenditure patterns, and local-level budgets. As an application, we examine the case of Portugal (1367–1844). Our findings demonstrate that throughout most of this extended period, Portugal maintained comparatively high fiscal and legal capacities. This challenges claims that Portugal’s economic decline from the second half of the eighteenth century was due to low state capacity.

1. Introduction

State capacity is a proximate cause of why some societies have better economic performance than others (Besley and Persson 2011; Dincecco 2017). The insight that countries with states lacking the ability to collect resources and convert them into public goods fail to develop has been confirmed in a variety of historical studies.¹ Nevertheless, measuring state capacity for premodern economies is not trivial, especially when data limitations are severe. Our paper serves two main goals. First, it proposes new measures for comprehensively measuring historical state capacity, expanding the pathbreaking comparative work by Karaman and Pamuk (2010, 2013). Second, it illustrates the gains of the proposed methodologies by analyzing the case of Portugal.

Research on historical state capacity is typically based on shortcut methods in which the key variable is studied via a proxy: fiscal capacity, as represented by the volume of state revenues (Karaman and Pamuk 2010, 2013; Brandt et al., 2014). Leaving aside the issue of the deflators used for now, fiscal capacity does not necessarily translate into high state capacity for three reasons.

First, while aggregate revenues are an understandable choice, the composition of the revenues also matters. In fact, as Bonney (1999, p. 10) hints, states financed by resources raised primarily outside their territory had lower incentives for state-building in the homeland. In such a setting, the argument goes, states whose finances relied on tribute or colonial revenues rather than on taxes are marginally less prone to create genuine state capacity.

Secondly, and more importantly, the size of the budget is not a reliable proxy because expenditure can instigate very different outcomes: states can use revenues to build further palaces or appoint more judges. Historically, rising fiscal capacity is often explained by increasing military involvement. European military confrontation of the late seventeenth and eighteenth centuries led states to increase their revenues (O’Brien 1988; Brewer 1988; Parker 1996). There is, thus, no good reason to believe that this translated into better access to public goods or state services if most of the additional revenues were used to pay for luxuries such as palaces. In sum, high fiscal capacity does not necessarily imply high legal capacity. As such, expenditure patterns are essential for measuring actual state capacity.

Likewise, no historical analysis of state capacity can be made without close consideration of a typically neglected agent, which was a major provider of public goods: local governments. Being closer to the populations they purported to serve, local authorities, although not formally accountable, had to deal with tighter budget constraints and closer scrutiny. Despite the small size of their budgets, municipal governments were a considerable builder of state capacity in their role as providers of public goods. Given their higher level of accountability relative to the central state, municipal decisions in taxation and spending potentially enjoyed a higher degree of legitimacy.

Only after addressing these four issues—the non-coincidence of the measures of fiscal and state capacity, the composition of revenues, the pattern of the expenditure, and the role of local governments—is it possible to design a more comprehensive methodology for measuring and comparing historical state capacity. This paper addresses these issues by building more robust measures of state capacity, which have several advantages over the current state of the art. We concentrate on Portugal as a case study to illustrate the insights that can be gained from considering a set of measures that go beyond the conventional option of using only central government revenues per capita deflated by nominal wages, as in the pioneering work of Karaman and Pamuk (2013). Absent from the existing fiscal capacity comparisons (Karaman and Pamuk 2013), the case of Portugal illustrates some limitations of the currently existing methodologies. Our analysis of this case will also test the oft-repeated claim that Portugal was beset by low state capacity and, accordingly, condemned to a dismal economic performance (Justino 1986; Tilly, 1992, Hespanha 1994, 2013; Yun-Casalilla 2019). Allegedly, this problem stemmed from the Portuguese Crown’s dependence on colonial revenues and, consequently, lack of incentives to tax and, hence, build up its fiscal capacity, like in today’s oil-exporting countries (Tilly 1992, p. 21, 62).² Such statements, however, are still to be tested systematically using the full range of available sources over time.

In this paper we build a new historical database, including data on expenditure patterns and on the size and role of the municipal governments, from the earliest aggregate data to the mid-nineteenth century. We also include data on the share of imperial revenues. With this data, we show that the comparative evidence does not support the interpretation that Portugal (comparatively) lacked fiscal capacity until the nineteenth century. Likewise, we show that imperial revenues only mattered for state finances during some specific periods. Our analysis also indicates that legal capacity increased over time as expenditure patterns evolved towards the provision of public goods. Finally, we show that local governments had a complementary role in building up overall fiscal capacity, but their resources were more oriented towards the improvement of legal capacity. Since Portugal did not lack state capacity, the fundamental causes for its eventual development failure must lie elsewhere.

2. Measuring Leviathan

Weak states fail to create conditions for development.³ States needed revenues to provide public goods such as national defense and courts, which in turn contributed to internal stability and market development. The symbiotic relationship between fiscal capacity and growth laid out by Epstein (2000) indicates that “in Europe more power seems to have allowed central governments to promote economic change and market integration” (Bogart et al. (2010, p. 94).

In Western Europe, fiscal capacity substantially increased over the early modern period and, through it, this region markedly diverged from the rest of Eurasia, notably Poland, Russia, the Ottoman empire, and China (Brandt et al., 2014, pp. 70–1, Karaman and Pamuk 2010, 2013). In the mid-seventeenth century, Netherlands and England attained higher levels of fiscal capacity than the other Western European powers (except for the small Republic of Venice). Which of these paths did Portugal tread?

Cardoso and Lains (2010, p. 251) concluded that “[i]n the beginning of the nineteenth century Portugal … [had] a relatively weak state.” This claim was built on a long tradition that analyzed early modern Portuguese finances (Godinho 1978; Hespanha 1994; Silva 2004; Carrara 2011; Oliveira 2013; Hespanha 2013). These analyses saw as the major shortcomings of the fiscal system the rigidity of the tax structure, which had to be compensated by a structural reliance on the empire and on revenues from the country’s intense maritime trade (Hespanha 1994, p. 142–3; Silva 2004, p. 246). Relying on few, and contradictory, international comparisons (Silva 2004, p. 240; Hespanha 2013), the influential historian A. M. Hespanha went on to claim that Portuguese fiscal capacity was one of the lowest in Europe (Hespanha 2013, p. 29).⁴

Much of the literature considers that at the heart of the Portuguese state stood an overreliance on colonial revenues going back to the sixteenth century; its wealth constituted a disincentive to develop a functional tax system (Tilly 1992, p. 62; Yun 2019, p. 262). For example, Tilly (1992, pp. 124–5) writes that “Spain and Portugal escaped the civilization of government by drawing on colonial revenues for a major share of military expenditures … Spain and Portugal anticipated … the situations of many contemporary Third World states in which military men hold power.” Likewise, Yun-Casalilla (2019, pp. 36, 48, 261–2, 401), writing about Portugal until the early seventeenth century, claims that vast imperial revenues took away incentives for developing a tax state.

In this paper we expand the analysis of Portuguese state revenues to include a wider time span from the earliest reliable quantitative data (Henriques 2009) to the period in which the liberal regime struggled to create a new fiscal system (Cardoso and Lains 2010). Like our predecessors, we cannot draw from complete series, but rather from a few benchmark years for which there is a comprehensive account of the revenues. The type of sources that we use for the numerator, concerning central government revenues, is illustrated in figure 1. After a considerable effort in source collection and source criticism, it was possible to include a few additional data points from the early modern period, unknown by the authors mentioned. We also distinguished between imperial and non-imperial revenues and include this distinction in our analysis.⁵

In table 1, we compare Portuguese fiscal capacity since 1369 with the now-familiar measure developed by Karaman and Pamuk (2013): day wages for unskilled workers.⁶ We also expand their dataset for the period 1350–1500, for England and France, for which good data exist,⁷ as well as Portugal. Deflated by per capita wages, Portuguese fiscal capacity appears comparatively high. If we refer to the 1750–1799 period, when the Little Divergence was already well advanced and for which we can compare the whole panel, Portugal had a seemingly robust comparative fiscal capacity.⁸ In the eighteenth century, imperial revenues allowed for a fiscal capacity inferior only to the Dutch Republic. Even without the empire, Portugal’s fiscal capacity would have been on the level of Spain, France, and Austria and well above the Eastern empires.⁹ The Portuguese high fiscal capacity was not, like Prussia for instance, the result of a steep rise in overall revenues. Instead, the pre-1500 observations show that this was a structural feature of the country since the fourteenth century, when the country became a ‘tax state’ (Henriques 2014), following the taxonomy created by Bonney and Ormrod (1999).

Table 1 illustrates, however, some limitations of using the nominal wage to deflate revenues. England’s fiscal capacity, in particular, appears lower in the second half of the seventeenth century, when parliamentary legitimacy allowed for increased volume and scope of taxation, and economic growth and structural change occurred (Broadberry et al., 2015; Wallis et al., 2018; Humphries and Weisdorf 2019; O’Brien 1988), as occurred in the Dutch Republic.¹⁰ England’s relatively high—and growing—wages excessively deflate revenues.¹¹ This suggests the need to consider alternative measures. The well-known “respectability basket” (Allen 2001), as adjusted to local consumption patterns (e.g., Palma and Reis 2019), is not directly affected by forces specific to each country’s labor market and, as such, provides a useful alternative. The comparisons using this alternative as deflator are shown in table 2.

As confirmed by the measures in table 2, Portuguese fiscal capacity increased from the earliest available observations, in line with the remaining Western powers. In the pre-1500 period, the nascent Portuguese tax state was not as large as England’s, while it was fighting the Hundred Years War, but it overshadowed that of France and continued to do so until the nineteenth century (except for the warlike and inflation-plagued 1650–1699 period). At any rate, the two tables agree that Portuguese fiscal capacity was well within Western European levels. The empire did contribute to reinforce Portuguese finances but, even if imperial revenues faltered, fiscal capacity would be at reasonable levels.

Analyses based on commodity baskets also have some shortcomings, as these are affected by relative prices that can lead to underestimating the fiscal capacity in the most advanced economies. Thus, in table 3, we consider the nominal revenues deflated by nominal GDP. Although this is the standard measure for present-day analysis, data for historical periods is harder to come by, as most of the reconstructed national accounts aim at estimating real per capita GDP rather than nominal GDP. The data starts in the 1500s, given that the reconstruction of Portugal's GDP for the period prior to the 16^th century is lacking.

Table 3 shows that Portuguese fiscal capacity was not far from the European leaders until the second half of the eighteenth century. This is a result consistent with the qualitative historical evidence (e.g., Brewer 1988; O’Brien 1988). Portugal’s central state revenues are also roughly similar with those of Spain and even France until the second half of the eighteenth century.¹² Overall, hence, the viewpoint that Portugal had a relatively high fiscal capacity until at least the early nineteenth century is robust to all three measurements employed here. In the next section, we show further details concerning the extent to which Portugal’s relatively high levels of fiscal capacity relied on its large overseas empire.

3. The empire: boon or burden?

The breakdown of imperial versus non-imperial revenues shows that Portuguese relatively high levels of fiscal capacity are not solely explained by the contribution of the empire. Imperial revenues were, nonetheless, important to keep up with England, France, and Spain. Table 3 shows that without imperial revenues, Portuguese fiscal capacity would have been considerably lower than that of England. The provocative sobriquet that Francis I of France chose for Portugal’s Manuel I (the roi épicier, i.e., the “spicer king”; Subrahmanyam 2012) was not entirely misguided as crown revenues of a colonial nature—especially the pepper trade—were a major contributor during the first half of the sixteenth century, as figure 2 shows. As mentioned, this might have led historians to consider that the overseas possessions underpinned Portuguese government revenues. And yet, this was not the case.¹³

It is also worth noting that the weight of imperial revenues over total central revenues declined over time: from more than 50 percent in the first half of the sixteenth century to 15 percent by the early eighteenth century (see figure 2). The share increased to around 25 percent in the middle decades of the eighteenth century, but then decreased again to 10–15 percent by the early nineteenth century. This evolution contradicts the claim that the imperial expansion arrested the development of the domestic fiscal system (Tilly 1992; Amaral 2012).¹⁴ Permanent, kingdom-wide, universal taxes like the sisas (from 1387 onwards) and the décima (from 1641 onwards) testify that the central government was able to create a tax state (Silva 2004; Henriques 2014), not to mention the increasing fiscal pressure under foreign rule (1580–1640). It should be noted that the large relative importance of imperial revenues with respect to total central state revenues in the sixteenth century was not indicative of the total size of overseas trade over time, which was only large relative to the size of Portugal’s economy during the eighteenth century (Costa et al., 2015).

The fiscal capacity of Portugal did not entirely rest on colonial revenues and monopolies. Since the late fourteenth century, the Portuguese state obtained three-quarters of its revenue from domestic sales taxes (Henriques 2009). Even though this situation changed briefly during the sixteenth century, when colonial revenues crept in, and for a more prolonged period over the eighteenth century, a strong dependence on colonial revenues cannot be considered the default situation of Portugal over the time span covered here.

In figure 3, we additionally consider Portugal’s custom revenues as a share of the kingdom’s revenues in the long run (1367–1810). This distinction matters because, like the empire, their major volumes may be regarded as a strong disincentive to develop new taxes and effective institutions. This influential argument was first proposed by Godinho (1978) and explored by Hespanha (1994, p. 142–3) and Cardoso and Lains (2010, p. 266) in their analyses.

As figure 3 shows, Godinho (1978) and others were right to highlight the role of the customs. However, their weight was only decisive in two specific periods: the first halves of the seventeenth and nineteenth centuries. When state finances were reliant on a new tax, like the sisa (since 1387) and the décima (since 1641), the share of the contribution of customs decreased. In other words, customs towered over the remaining sources only in periods when the central state did not tap domestic economic activity with a well-designed tax. As such, the funds awarded by the country’s burgeoning trade did not deter the emergence of new taxes.

Overall, the evidence analyzed in these two sections shows that Portuguese fiscal capacity was not a factor in the dismal performance of the Portuguese economy since the late eighteenth century (Palma and Reis 2019).

4. Expenditure and legal capacity

“We all ought to pay taxes because the king undertakes the defense of the homeland and protects us from enemies and thieves, maintains the peace and the justice among men, as well as [he] builds and repairs fortresses,”—thus wrote the chaplain of King Manuel I in his 1496 treatise (Rebelo 1951). While this was an ideal portrait rather than an exact reading of the expenditure side of the budget, it reflects the widely held assumption that taxes served common purposes rather than the personal expenses of the monarch. Contemporary political society understood this: answering criticism of the municipalities about the sales taxes in the 1498 Cortes, Manuel I provocatively replied that “if we ceased to collect them, you would later beg me to be impose them again” (see Henriques 2009, p. 264). The king thus hinted that the public was the ultimate beneficiary of fiscal capacity.

The analysis of expenditures provides a window into what Besley and Persson (2011, p. 6) call legal capacity, that is the ability to convert fiscal capacity into expenses that raise private sector productivity, like public works, regulation, and law enforcement. Rulers who spend revenues building palaces or engaging in negative-sum dynastic wars are not using tax funds to provide for public goods and the promotion of economic growth. With their revenues, rulers could fund public goods—such as roads—or their own conspicuous consumption, such as palaces. In other words, fiscal capacity measures concern inputs, not outputs or their efficiency. Hence, a comparative analysis of the concentration of resources available to the state does not directly translate into the provision of public goods as commonly understood.

In this section we break down the structure of central public expenditures (figure 4). For the analysis, it is paramount to obtain comparable categories of expenditure that can be estimated with the accounting system found in the available sources. We follow what is called a functional analysis, i.e., we use categories that can be aggregated by adding the respective state agencies. This allowed us to disaggregate the expenditure into meaningful categories. Two categories of expenses, royal household and redistribution, broadly (but not totally) indicate whether revenues were diverted to a narrow group of interests (including the monarch) or, instead, to productivity-enhancing works and services. The royal household covered expenses with the maintenance of the monarch, his family, and his court, including scions of noble families in their teens brought up who would provide many of the military and administrative positions. As the young men were being brought up, the state budget accounted for their moradias, i.e., living expenses as “dwellers” (moradores) of the royal household. The moradias were only a share of the royal household: in 1527 and 1543, they reached 31 and 35 percent, but in 1563, represented only 5 percent of the expenditure categorized as Royal Household. Redistributive payments consisted of subsidies to the households of the royal family and of great nobles, pensions, and scholarships as well as dowries.

Redistribution payments rested on two payments: assentamentos, given to the large households of great nobles including royal family members, and tenças, which rewarded services to the king (scholarships and dowries were only important in the fifteenth century). The typical recipient of a tença was a nobleman who had served without pay for years in the hope of a royal reward (see figure 1). In turn, the expenses that we categorize as military are different, as they involve pay and maintenance of soldiers, as well as matériel, including expensive fortifications and ships.¹⁵ Its volume is determined by the current, not past, military commitments. In the eighteenth century, military expenditure rose as the consequence of the deliberate creation of a professional standing army, with a given number of regiments and fortifications, routine military promotions, scheduled careers, like it was becoming in contemporary Europe (Korner 1995). Ensuring the military capacity of a standing army meant a fixed cost structure. In sum, the military and redistribution categories do not overlap.

Other public goods encompass administration, the judicial system, and public works (and some modest education expenditure also appears from 1800). These expenses translate to legal capacity. Military expenditure is separated from the remaining public goods, given that it might be used to build up archetypal public goods (the defense of the realm) or to engage in negative-sum wars. At the last, debt service is included because it limited the fiscal space for improving legal capacity.

Figure 4 allows for distinguishing two legal capacity regimes. The earlier regime is first apparent in the 1473 budget (though it likely started earlier) and lasted for about one century. It covers much of the sixteenth century, when imperial revenues were important (despite the small size of intercontinental trade in relation to the overall economy at the time; Costa et al., 2015). In this regime, Portuguese expenditure was dominated by redistribution and by the costs of the royal household. Until the second half of the seventeenth century, these two categories reflected the redistributive means whereby the Crown attributed resources to maintain and reward a militarized elite (economia da mercê).¹⁶ These two categories often constituted about 50 percent of central expenses.¹⁷ While redistributive expenses allowed for the building of an empire, they ate away fiscal space. Military expenses became the dominant share of expenditure, as it was the rule in Western Europe (Korner 1995; Hoffman 2015). In this second regime, the share of expenditure assigned to public goods (mainly courts and administration) was kept above the 10 percent mark and increased throughout the period.

Domestic military weakness, among other factors, proved fatal for Portuguese hopes of retaining independence in 1580. The loss of independence in that year led to the emergence of a very different expenditure pattern. When the country was ruled by Spanish kings (1580–1640), there was no royal court in Portugal. Hence, court expenses were dramatically reduced, whereas the linkage between the vassals’ military service and the rewards given by kings weakened considerably. Thus, the weight of “royal household” and “redistribution” categories fell dramatically and would not be recovered when they resurfaced in the second half of the seventeenth century. Essentially, the share of expenses once occupied by redistribution and royal household was diverted to direct military expenditure (like in contemporary Prussia, Austria, France, and England), but with modest debt service, unlike the latter (Costa and Miranda 2023). This observation is consistent with the comparatively high level of fiscal revenue observed in section 2.

With this analysis we can see that Portuguese fiscal capacity —in the sense of ability to collect revenues— was effectively converted into military and legal capacity from the late sixteenth century onwards, and empire revenues did not hinder this process. The displacement of courtly and redistributive payments by making room for military and war-related debt repayments was also seen in the finances of other absolutist states such as Austria and Prussia during the eighteenth century (Korner n.d.; Dickson 1987, pp. 385–6). On the other hand, from 1645 to 1815, Portugal avoided the rapid increase of debt service and arrears that affected European states, where expenditure was related to war and public debt servicing (Alfani and Di Tullio 2019, p.172). In Portugal, the modest rise of debt service and arrears (until the nineteenth-century civil wars) allowed for increasing the portion of the budget dedicated to maintaining and erecting public goods and services. This share reached 13 percent in 1588 and was kept at that level until it approached 20 percent in the 1840s. This considerable share of the fiscal resources devoted to non-military public goods does not provide the whole picture. As it was common throughout early modern Europe, Portugal was a devolved state in which such expenses were left to the self-ruling communities that covered the entire territory: the municipalities.

5. The local level: too small to fail

Central state revenue and expenditure is only a part of state capacity. Local administrators also collected taxes and delivered public goods, and they worked independently of the central state. Since the twelfth century in Portugal, this role was played by some 800 municipalities, which administered nearly all the territory of the country.¹⁸ These authorities not only guaranteed public order and social cohesion as the building and maintenance of infrastructures were largely left to their resources. Likewise, the regulation of local markets, licenses, weights and measures, trial of minor offences, and public ceremonies and festivities was supported by the municipal budget. Transfers from the central government to the municipalities were minimal and were typically assigned to strategic, national goals. The collaboration of the central state was occasionally required in a few specific instances in which very large structures, like aqueducts or fortifications, had to be built.

Unlike the central state, which acted with little accountability and were vulnerable to political pressures, municipalities were “too small to fail”. Stricter budget constraints, closer accountability by the central state and the people alike, and better alignment with the interests of the taxpayers all implied that municipal resources contributed disproportionally to the provision of non-military public goods. As the people of Coimbra wrote in a 1459 petition, taking part in the municipal meetings was important to them because they felt the burden of paying [taxes] and “wanted to know how the affairs went and how municipal monies was spent” (Oliveira 2010, p. 77, our translation).

Tighter accountability and proximity meant that local authorities enjoyed the legitimacy and the capacity to foster social cohesion. In this setting, the central government devolution of such local-level expenses and its focus on public goods that had non-rival and non-excludable characteristics at the national level (e.g., national defense) represented a more efficient arrangement than the alternatives.

In this largely devolved system, the role of the municipalities was complementary to the central state. Despite their administrative and judicial importance, the mainstream historiography downplays the fiscal capacity of the municipalities, with the arguments that they did not mobilize significant resources and/or their expenditure was focused on personnel (e.g., Oliveira 1996, p. 131; Monteiro 1996, p. 135). This view is built on political discourse rather than on quantitative evidence. By aggregating a representative sample of municipalities over a long period, we show here that this judgment severely understates the role of the local level. While municipal budgets were small when considered individually, aggregate municipal revenues enhanced central fiscal capacity by about an additional 7 percent over the period (see table 4). More importantly, municipalities ensured about one-third of all expenditure in public goods over the period (see table 7).

5.1. Local revenues

Lacking the power to tax income or wealth, local governments tapped the areas of the economy that were left out by the central state. These marginal sources of revenue varied across the municipalities, but most of the municipalities depended on three sources: fines on economic activity, tolls and indirect taxes, and farming of the commons and natural resources. All municipalities collected a great variety of fines for minor offences (trespasses of cattle, public disturbances, failure to comply with regulations on quality standards or weights and measures) and charged fees for conducting inspections and awarding licenses. Municipal governments collected tolls and taxed certain professional activities and the consumption of some products. Practically all municipalities laid claim to the commons and hence charged their users, whilst the larger municipalities were also endowed with agrarian land and houses, whence they collected rents (for a detailed analysis, see Barbosa 2020).

Out of the total revenues, one-third (known as “terças”) was pooled at the regional level and managed by a centrally appointed guardian to sustain the building and repair of fortresses or naval defenses. These were not understood as transfers to the central state revenues, because the funds of the terças had to be spent on defensive works or, with royal license, on repairs of almshouses and hospitals. There were, however exemptions, as some municipalities were relieved of its payment during times of war, and there is at least one important municipality (Coimbra) that was exempted from it (Barbosa et al., 2022). The values of the tax farms of the terça for the whole kingdom in the period 1621–1688 (Oliveira 2013, p. 325) represent about one-quarter of the total municipal revenues estimated. There were also transfers from state taxation to the municipalities, but these were modest.¹⁹

To estimate the overall contribution of local revenues to Portuguese fiscal capacity, we aggregate dispersed local revenue information and measure it with respect to the revenues of the central state. We gathered monographs, published sources, and manuscript sources for twenty-four municipalities (see Appendix C) spread throughout the country (see figure 5). Our choice of this sample was based on all the available published information plus additional archival work that we did to ensure the geographic and chronological balance of the sample. The results of this effort can be seen in table 4. We find that the level of local revenues relative to central revenues was relatively stable. Except for two periods of intense military effort at home—the Restoration Wars of 1640–1668 and the French Wars of 1792–1815—which led to the diversion of resources to the central treasury and the erosion of the tax base (Barbosa 2017), local revenues corresponded to about one-tenth of central revenues (averaging 8.2 percent for all years observed, with extreme values at 3.6 and 13.6 percent and a standard deviation of 3.6). The contribution of the local level to overall Portuguese fiscal capacity was small, as expected, but not irrelevant. The key issue, however, is how these funds were spent.

5.2. Local expenditure

Although some historians did endeavor to study municipal expenses, their analyses typically used their own classification schemes and are typically limited to patches of time, making it difficult to systematize and aggregate the results of their research. Also, the low organizational complexity of the municipalities does not allow for a functional analysis. The results of this prior research from a few of the municipalities across the country (see figure 5 for their geographical locations) and for a major city like Porto can be seen in tables 5 and 6, respectively.

Table 5 shows how the expenditure pattern in different municipalities in different periods exhibits two traits that set it apart from the central state. The first is the absence of redistributive expenditure and of expenses with the royal household. The second is the role of public goods: more than half the money was typically spent in military, justice, and administration expenses during the entire early modern period. While the expenditure on religious and civic celebrations might appear detrimental to common welfare, this is to ignore the literature that shows such events built up cohesion and stability (Harris 2017). Collective celebrations (civic and religious), which were paid for locally, contributed to social cohesion: they helped make political institutions consensual.²⁰ Also, like charity expenses, they were primarily directed to the poor. As observed in other European countries, social welfare redistribution was essentially entrusted to the local level (Alfani and Di Tullio 2019, p. 166).

The large share spent on officers’ salaries, social assistance, and popular or religious celebrations shows that local government was a central piece in social cohesion and order without much central interference. The administrative and judicial costs paid at the local level helped support the central state. Expenditure on officers’ salaries allowed for the regular functioning of local institutions, and thus, political stability. Porto, whose municipal accounts range from the fifteenth to the later eighteenth century, provides a less dispersed picture of this evolution (table 6).

While the data for Porto confirm the broad picture contained in table 5, it should be observed that Porto was a large city and its commercial, naval, and administrative importance justified transfers from the central state and for a higher role of military expenditure (Valente 2008, p. 82; Costa 2018, p. 282). The municipality of Coimbra provides the best documented case over a long period of time (Barbosa et al., 2022; Barbosa 2023) and was more representative of the situation of municipalities across the country, as it lacked Porto’s strategic value and size. Moreover, the sources from Coimbra allow us to isolate the critical expenditure variable: non-defense public goods. Table 7 shows per capita expenses in public goods in the familiar unit of unskilled day wages and as percentage of total expenditure.

According to table 7, throughout the period, local-level expenses typically contributed to at least one-third of the provision of public goods countrywide. The high share of local budgets committed to non-defense public goods compensated the small size of the municipal contribution to overall expenditure. Nevertheless, the role of municipalities as providers of public goods diminished as the central budget started to dwarf its local counterparts and, simultaneously, municipal expenses became less focused on public goods. This finding shows the surprising importance of local-level expenditures in assessing state capacity in these historical periods. The oversized contribution of the municipalities to public goods liberated the central state to effectively affirm the country’s sovereignty in a geopolitically delicate situation, at least until the Napoleonic Wars came.

In sum, tighter accountability and proximity meant that local authorities enjoyed the legitimacy and the capacity to foster social cohesion. In this setting, the central government devolution of such local-level expenses and its focus on public goods that had non-rival and non-excludable characteristics at the national level (e.g., national defense) represents a marginally more efficient arrangement than the alternatives.

6. Conclusion

The case study of Portugal shows the advantages to be gained from exploring a wider range of variables to measure state capacity. The central importance of state capacity to explain economic development requires economic historians to overcome the limitations of historical data and develop empirical methodologies for measuring state capacity. These measures must be comparable over long periods of time. In this paper we explored new methods, and we illustrated these using Portugal as a case study.

Overall, our new fiscal capacity measures paint a similar picture with respect to Portugal: the country had average Western European fiscal capacity levels. This capacity was not built on imperial revenues. With regards to the weight of the empire for fiscal revenues, Portugal went through three main regimes over the early modern period. In the sixteenth century, high imperial revenues mattered for the overall fiscal position of the Crown, despite the low overall importance of intercontinental trade relative to the size of the economy (Costa et al., 2015). In the eighteenth century, overseas revenues regained some of their prior importance, even though they never became dominant as a share of central revenues.

Prior to the mid-nineteenth century, Portugal did not lack state capacity. However measured, Portugal’s fiscal capacity rose over time, although it fell short of the European leader. More importantly, when central government provision of non-military public goods is added to the local level, we see an increase in the supply of public goods. Hence, and by contrast to much of the prior literature (e.g., Tilly 1992, Justino 1986, Hespanha 1994, 2013), we argue that low fiscal capacity was not the key factor in the country’s dismal economic performance. In the fifteenth and sixteenth centuries, there was a high level of associated redistribution and royal household expenditure. But by 1600, the weight of these expenses lost importance, and their fiscal space was occupied by military expenditure. Also, the country assigned a significant share of its expenditures to public goods other than defense. Overall, this implies that other factors must explain its comparative development failure from the eighteenth century onward (Palma and Reis 2019; Henriques and Palma 2023; Kedrosky and Palma 2024). Also, Portugal’s macroeconomic failings after the second half of the eighteenth century should not distract from the fact that its institutional fabric, at both the central and the local level, was able to convert fiscal resources into legal capacity. Also, we found that the arrangements between the local and the central level provided the country with cohesion, political stability, and average levels of fiscal capacity, even if these were not sufficient to cause long-run economic growth.

Acknowledgements

This paper previously circulated under the title “Portugal’s early modern state capacity: a comparative approach.” We are grateful for advice and comments from Guido Alfani, David Chilosi, Guillaume Daudin, Giovanni Federico, Marcelo Fernandes, Cristina Giesteira, Teresa Castro Henriques, Kivanç Karaman, Pedro Lains, Susana Münch Miranda, Sevket Pamuk, Pedro Pinto, Jaime Reis, Lisbeth Rodrigues, as well as participants in seminars and conferences at ICS-UL, University of Manchester, Groningen EHES, and Paris WEHC. We thank José Luís Barbosa for sharing data concerning the municipality of Coimbra, Leandro Prados de la Escosura for sharing Spain’s nominal GDP, Leonardo Ridolfi for sharing France’s nominal GDP, António Alves Caetano for sharing his expenditure estimates for 1815, and Oriol Sabaté for organizing the state capacity session at Paris WEHC. Joakim Book provided excellent research assistance. The usual disclaimer applies. The authors also acknowledge financial support from Fundação para a Ciência e a Tecnologia (CEECIND/04197/2017 and PTDC/HAR-HIS/28809/2017).

Data availability

The data and replication package for this article are available at: https://www.openicpsr.org/openicpsr/project/198143/

Footnotes

References