2. The literacy transition: a supply- or demand-driven process?

Leaving aside North America, the transition from restricted to universal literacy first took place in Northwestern Europe, hand-in-hand with the establishment of a system of formal schooling and growing economic development (Easterlin, Reference Easterlin1981; Núñez, Reference Núñez2003a; Sandberg, Reference Sandberg1982). Spain, however, experienced a delayed and geographically uneven spread of literacy (Núñez, Reference Núñez1992, Reference Núñez2003b). Although growing, Spanish educational attainments diverged from the levels achieved by the most developed nations during the second half of the 19th century and only began to converge during the first decades of the 20th century. These improvements were also subject to frequent interruptions and setbacks, especially between 1870 and 1910 (Núñez, Reference Núñez and Morilla2010: 253). Likewise, regional differences on literacy rates, already high in 1860, widened during the second half of the 19th century and the first decades of the 20th century (see Figure 1). Even though educational attainments were improving everywhere, the early stages of economic modernisation resulted in a huge educational gap between Northern and Southern Spain.Footnote ⁵

Figure 1.

The literacy transition in Spain (percentage of population literate). Source: Núñez (Reference Núñez1992).

The demand for education has been considered as a crucial element in the rise of human capital during this period. Only when the returns to education were large enough were investments made to acquire it (Galor and Weil, Reference Galor and Weil1999; Núñez, Reference Núñez2003a; Reis, Reference Reis, Allen, Bengtsson and Dribe2005). The rate of return to education, and thus the incentives to invest in it, depends on its potential benefits and actual costs. In this sense, apart from the direct costs of schooling, the opportunity cost of forgoing work is crucial in household behaviour. Literacy and primary schooling had a higher economic value for the average individual in more market-oriented economies and in the growing non-agricultural sector (Lucas, Reference Lucas2002; Sandberg, Reference Sandberg1982). More developed societies also had more resources to invest in education. The relationship between industrialisation and literacy, however, is not so clear. Although some scholars argue that literacy was related to the British industrial revolution, other researchers have pointed out to stagnating, or even falling, literacy rates during that period, especially in industrial areas (Mitch, Reference Mitch1992; Schofield, Reference Schofield1973).Footnote ⁶ According to Galor (Reference Galor2011), the acceleration of technological progress during the second phase of industrialisation gradually increased the relative importance of human capital. Likewise, an unequal access to land may have also influenced the demand for education. The existence of credit market imperfections disproportionately affects the lower classes’ capacity to invest in human capital (Galor and Zeira, Reference Galor and Zeira1993). Moreover, while small and middle size farmers positively valued education, landless labourers, due to scanty economic prospects, did not see any economic advantage from investing on it (Núñez, Reference Núñez, Jerneck, Mörner, Tortella and Åkerman2005: 132).Footnote ⁷ Since the decision to invest in education is generally taken by the parents, their educational levels, particularly those of the mothers, significantly influenced the demand for education for the next generation (Pérez Moreda, Reference Pérez Moreda1997). Lastly, life expectancy, by determining the time horizon of the investment, also affects the potential benefits that can be reaped off from investing in education (Galor, Reference Galor2011: 52; Núñez, Reference Núñez, Jerneck, Mörner, Tortella and Åkerman2005: 131).

Although demand factors are crucial, the supply side, by affecting the costs of education, plays an important role in determining the levels of human capital (Lindert, Reference Lindert2003; Núñez, Reference Núñez2003a; Reis, Reference Reis, Allen, Bengtsson and Dribe2005). The public provision of free and compulsory primary schooling significantly reduced the direct cost of education, while at the same time it tried to prevent child labour. The first serious attempt to promote primary schooling in Spain was the Moyano Act in 1857 (Núñez, Reference Núñez1992: 216–226).Footnote ⁸ However, although it theoretically established compulsory education (and free for those who could not afford it), its enactment was not fully effective until the early 20th century. Recent research has shown that an unequal distribution of land property rights negatively influences human capital formation (Deininger and Squire, Reference Deininger and Squire1998; Engerman and Sokoloff, Reference Engerman and Sokoloff2000; Galor et al., Reference Galor, Moav and Vollrath2009). Landholding elites are likely to block the implementation of educational reforms in order to both preserve the power relations status quo and reduce the mobility of the rural labour force. Although the backwardness of the economy obviously limited the spending capacity during this period, the low levels of public expenditure on schooling evidenced the lack of commitment of the Spanish government and local elites to primary education, partly explaining the deficient primary educational system (Comín, Reference Comín1988; Núñez, Reference Núñez1991, Reference Núñez, Jerneck, Mörner, Tortella and Åkerman2005). The finance of schooling was actually left to municipal councils and, therefore, their capacity to fund primary education became crucial. Núñez (Reference Núñez1991) has shown that schooling expenditures per capita were already geographically uneven in 1860 and these regional differences had increased in 1930. A related problem was the low quality of school teaching. The ratio students per teacher worsened throughout the 19th century, probably up to around 1910 (Núñez, Reference Núñez1992: 237). The material conditions of the schools were deficient and teachers’ low salaries and social status prevented attracting qualified and motivated professionals (Núñez, Reference Núñez, Jerneck, Mörner, Tortella and Åkerman2005: 130). Although it soon became obvious that municipalities could not afford supporting a well-functioning primary education system, the central government was too slow to intervene: it was only in 1902 when, together with the creation of the Ministry of Public Instruction and Arts, it took care of the salaries of school teachers (Núñez, Reference Núñez1991). However, given that the central state limited itself to cover schooling expenditures, without readjusting the pre-existing disequilibria, the problems regarding the uneven regional distribution of schooling expenditures persisted well into the first decades of the 20th century. Furthermore, the problem with primary education was not only of lack of resources but also of allocation between educational levels. Those regions where elites were powerful enough gave priority to secondary and university level expenditures, thus negatively affecting primary schooling (Núñez, Reference Núñez1991).Footnote ⁹

However, apart from government intervention, other elements should also be taken into account when explaining the supply of education. Although demographic growth in Spain was relatively slow, the proportion of schooling-age population was significantly larger than in other countries (Núñez, Reference Núñez, Jerneck, Mörner, Tortella and Åkerman2005: 131). This trend put an extra pressure on the Spanish educational system, making it more difficult to increase schooling expenditures per capita.Footnote ¹⁰ It may be the case, nonetheless, that the supply of education enjoyed increasing returns to scale, so the extra demographic burden may have been offset by the increasing productivity arising from size (Pérez Moreda, Reference Pérez Moreda1997: 249). Likewise, although distance to the school is likely to have played a negative influence in areas with dispersed settlements, especially in the context of a steep orography, the one-teacher per village policy, related also to the role that the local parish had on imparting education, may have benefited small villages (Pérez Moreda, Reference Pérez Moreda1997: 249; Sarasúa, Reference Sarasúa and Martínez-Carrión2002: 569–570). Lastly, inherited historical factors, such as somewhat different legal systems or cultural values, may have also played a role in the different regional paths followed by educational attainments in Spain during this period (Pérez Moreda, Reference Pérez Moreda1997; Reher, Reference Reher1997).

3. Commons and human capital

Commons have been traditionally regarded as inefficient and privatisation has been usually regarded as a precondition to foster economic growth (North and Thomas, Reference North and Thomas1977). The negative view surrounding the communal regime has nonetheless been subjected to revision by a new wave of empirical research that considers common property regimes to be efficient and sustainable, thus revaluating the role that common resources had for the local communities that managed them (Allen, Reference Allen1992; De Moor, Reference De Moor2009; De Moor et al., Reference De Moor, Shaw-Taylor and Warde2002; Ostrom, Reference Ostrom1990). The reassessment of the commons has focused on their impact on agricultural productivity and inequality (Allen, Reference Allen1992, Reference Allen2003; Humphries, Reference Humphries1990; Neeson, Reference Neeson1993), but their potential contribution to human capital has hardly been explored.

Common lands were a key component in the organic-based Spanish preindustrial economy (Balboa, Reference Balboa1999; Iriarte, Reference Iriarte2002; Jiménez Blanco, Reference Jiménez Blanco2002). The communal regime in Spain involved two main types of access to the land: a direct but regulated access for all members of the community (bienes comunales) or a temporary cession of user rights to particular individuals in exchange for a monetary income (bienes de propios). Apart from providing pasture to support livestock, which in turn supplied agriculture with fertiliser and workforce, these comunales constituted a source of complementary income by providing animal proteins, wood, and fuel, among other products, including the possibility of temporary cropping. The bienes de propios, on the other hand, played a fundamental role in the finances of local institutions, which was particularly important given that municipalities were responsible for the provision of basic public services. However, the transformations caused by the transition to capitalism, and the emergence of a new liberal state, triggered the gradual dismantling of the communal regime throughout the 19th and early 20th centuries (Iriarte, Reference Iriarte2002). The degree of common land persistence was nonetheless fairly different depending on the region being analysed (GEHR,Footnote ¹¹1994). As shown in Figure 2, the dismantling of the communal regime was particularly intense in the half south of the country, while common land persistence was especially high in North-western Spain.Footnote ¹²

Figure 2.

Common land persistence in Spain (percentage of total area). Source: Artiaga and Balboa (Reference Artiaga and Balboa1992), GEHR (1994) and Gallego (Reference Gallego2007). No data for the Basque Country are available.

During the 19th century, municipalities were indeed responsible for the provision of elementary education and their financial capacity was crucial when it came to funding schooling expenditures.Footnote ¹³ It has been argued that both the backwardness of Spanish literacy and its uneven regional spread are partly explained by the fiscal problems of these local institutions (García and Comín, Reference García and Comín1995; Núñez, Reference Núñez1991). The monetary income derived from the cession of user rights on the commons actually constituted a fundamental component of the municipal budget (Bernal, Reference Bernal1978; García and Comín, Reference García and Comín1995; Iriarte, Reference Iriarte, Sebastián and Uriarte2003; Linares, Reference Linares2006).Footnote ¹⁴ In 1858, revenues obtained from the commons met 32.4% of the municipal budget (García and Comín, Reference García and Comín1995: 95).Footnote ¹⁵ It should be noted that these figures reflect the national average and therefore hide the importance of the commons in those municipalities that had preserved them, especially in the rural areas. In the province of Seville, for instance, despite being one of the areas that most suffered privatisation prior to the Disentailment Act of 1855, the income generated by the commons still provided the 100% of the ordinary revenue in 66% of the municipalities in 1849 (Bernal, Reference Bernal1978: 307).Footnote ¹⁶

The financial difficulties of municipalities during the 19th century are well known by the historiography (García and Comín, Reference García and Comín1995; Moral Ruíz, Reference Moral Ruíz and Alonso1986). The privatisation of common lands meant a loss of assets that negatively influenced their economic viability and the possibility to meet the increase in expenditures required by the functions on education, among other basic public services, which they were suppose to carry out (Iriarte, Reference Iriarte, Sebastián and Uriarte2003; Jiménez Blanco, Reference Jiménez Blanco2002: 169; Linares, Reference Linares2006).Footnote ¹⁷ In Seville, for instance, the revenues generated by common lands were reduced by 70% between 1821 and 1849, a situation that aggravated later as a result of the General Disentailment Act of 1855 (Bernal, Reference Bernal1978: 302).Footnote ¹⁸ The financial problems of municipalities negatively influenced the provision of local public goods, especially of education (García and Comín, Reference García and Comín1995). Most of the municipal budget on education was devoted to the payment of local teachers, which did not prevent their wages from being extremely low. In addition, the delays in paying the wages of the local teachers were widespread. A high proportion of these teachers had no official certificates and a great deal of villages did not have an adequate building devoted to the school.

Although literacy improved in the whole country during the period analysed here, the gap between Northern and Southern Spain increased (Núñez, Reference Núñez1992). The provision of schooling, measured by the number of schools and teacher per population, and the public expenditures in education per capita, were indeed higher in the northern half of the Peninsula, which also coincided with the geography of common land persistence. The diverse survival of the commons may therefore partly contribute to explaining the dissimilar funding capacity of municipalities that led to a considerable regional variation in the provision of schooling (Collantes, Reference Collantes2004; Núñez, Reference Núñez1992; Sarasúa, Reference Sarasúa and Martínez-Carrión2002). In the province of Cádiz, for instance, most of the teachers’ salaries came from revenues generated by the commons in 1840 (Bernal, Reference Bernal1978: 303). Furthermore, municipalities frequently financed schooling directly through the commons by allocating a plot of land to the maintenance of the teacher or by providing the building where lessons were given (Sarasúa, Reference Sarasúa and Martínez-Carrión2002: 580–581).Footnote ¹⁹

However, apart from contributing to the supply of education by financing public services, the income generated by the commons allowed the reduction of the fiscal burden supported by the community (García and Comín, Reference García and Comín1995; Iriarte, Reference Iriarte, Sebastián and Uriarte2003; Linares, Reference Linares2006). The privatisation of these collective resources coincided with the increasing expenditures that municipalities had to face to provide new public services. These two combined factors not only influenced the provision of education and other public goods, such as medical care and poor relief, but also forced local institutions to increase taxes. Poorer households were especially affected by this process due to the regressive nature of a fiscal system built mostly around taxing consumption goods.Footnote ²⁰ This consequence was by no means unexpected for the contemporaries. The parliamentary debates about the convenience of privatising common lands carried out between 1835 and 1855 reflect the concern that depriving local communities from these resources would necessarily force municipalities to increase local taxes (Gómez Urdañez, Reference Gómez Urdañez2002: 144). Municipal budgetary problems also meant that local public goods, especially schooling, were sometimes funded through neighbours’ extraordinary contributions (Sarasúa, Reference Sarasúa and Martínez-Carrión2002: 581).Footnote ²¹

Likewise, the dismantling of common lands not only influenced the municipal financial capacity and the level of local taxes, but it also directly affected households’ incomes. As explained above, common lands provided pasture, fertiliser, wood, fuel, among other products, including the possibility of temporary cropping. The loss of these sources of complementary incomes, by reducing disposable income, increased the relative cost of education, thus reducing the demand for education, particularly for the less favoured groups. The widespread conflict and resistance that privatisation generated strongly points to the crucial role that commons played on securing the subsistence of rural households and the negative impact that privatisation had on their living standards (Cobo et al., Reference Cobo, Cruz and González de Molina1992; De la Torre and Lana, Reference De la Torre and Lana2000).

Lastly, the privatisation of common lands could also have indirectly influenced human capital through its effect on inequality. The way through which privatisation was implemented is likely to have increased, or at least consolidated, the concentration of landholding by an elite, thus contributing to social polarisation and the proletarisation of agricultural labour (Linares, Reference Linares2001; Moral Ruíz, Reference Moral Ruíz1979). According to the arguments outlined in Section 3, an unequal distribution of land property rights negatively affected human capital formation through both the supply and the demand for education. Not only did landless labourers not see any economic advantage from investing on it, but also landholding elites blocked the implementation of educational policies.

6. The literacy transition: economic modernisation or state intervention?

Interestingly, the results reported in Table 1 also allow assessing the relative importance of these two processes, economic modernisation and state intervention, in the transition to universal literacy in Spain. The coefficients of the time-period fixed effects illustrate that as the country developed, both indicators of human capital greatly improved, especially during the first decades of the 20th century. If we compare the coefficients of the time dummies before and after including the host of controls, the distinctive impact of the modernisation process and the increasing role of the state can be distinguished. In columns (2) and (6), the time dummies capture the combined impact of both processes. Literacy rates increased throughout the whole period, although the increase is almost two times larger between 1900 and 1930 than during the previous period. Schooling expenditures grew evenly during the whole period. Columns (3) and (7) add the set of controls reflecting the ongoing modernisation process visible in growing incomes, urbanisation and industrialisation levels, together with other demographic and economic factors. The coefficient on the dummy variable now reflects the rise in literacy and schooling expenditure that is not explained by the model and can therefore be attributed to the efforts of the state to improve the provision of education. The comparison between these pairs of columns unveils interesting conclusions.

On the one hand, regarding literacy rates, the effect of the dummy for 1900 hardly changes when controls are included, which means that economic modernisation barely had any effect on human capital and that most of those 17 percentage points increase on literacy rates between 1860 and 1900 were due to state intervention. This situation changed during the first decades of the 20th century. Although the state was still responsible for a larger share of the now higher increases in literacy rates, economic modernisation began to play a significant role in pushing human capital forward.Footnote ³⁹ On the other hand, although schooling expenditures per capita grew evenly during both periods, the results evidence that the process of structural change triggered off by economic modernisation had negative consequences on the supply of primary education. The coefficient of the time dummies increases significantly after including the set of controls, implying that the financial effort of municipalities and the state was partly offset by other negative processes at play. In this sense, accelerating population and urbanisation growth is likely to have put more pressure on the scarce resources available (Lindert, Reference Lindert2003: 343; Núñez, Reference Núñez and Morilla2010: 256).

These results therefore strongly confirm the crucial role of public institutions in providing education, especially in the first stages of economic development. It is worth noting that the contribution of the public sector to literacy rates and schooling expenditure remained roughly unchanged both between 1860 and 1900 and between 1900 and 1930.Footnote ⁴⁰ This points to the robustness of these results since the role of the state in providing education, at the local or national level, is reflected first on schooling expenditures, which is then translated into literacy rates, a measure that reflects the interaction of supply and demand factors. Although there is evidence of some improvements, the modernisation of the Spanish economy was not enough to foster an endogenous increase in human capital indicators. Economic development only translated into a noticeable increase in literacy rates from 1900 onwards and, even then, its contribution was significantly lower than that of the public sector. Therefore, either demand forces were too weak or their role in promoting educational human capital was not as important as it has been suggested.

However, given the low levels of human capital in Spain relative to neighbouring countries, it can be argued that the public effort was not enough either. Likewise, given that the contribution of the public sector to both literacy levels and school expenditure is similar throughout the whole period, it seems that the Law of 1902, by which the central government took care of teachers’ salaries, thus releasing municipalities from that responsibility, did not significantly succeed in promoting human capital indicators. In this sense, Núñez (Reference Núñez1991: 145) argues that the centralisation of schooling expenditure was not able to alter previous trends because it limited itself to meeting the pre-existing budget without either improving it or adjusting its uneven regional distribution.Footnote ⁴¹ According to this author, an oversized secondary and tertiary educational sector, mostly benefiting the sons of the elites, prevented more public resources to be devoted to schooling. In this sense, while there was a wide gap between Spain and other European countries in terms of investment on primary schooling, the difference is hardly significant in secondary education and even positive at the university level (Núñez, Reference Núñez and Morilla2010: 247–252). The distribution of public expenditures among the different educational levels was thus not only unequal but also inefficient given the importance of primary schooling for economic growth in developing countries (261–264).

Several authors have pointed to the lack of interest of the elites to educate the masses, reflecting the high degree of inequality in the Spanish economic and political arena (Núñez, Reference Núñez1992; Pérez Moreda, Reference Pérez Moreda1997; Reher, Reference Reher1997). In this regard, Easterlin (Reference Easterlin1981: 14) claims that a decisive commitment to mass education only happens when ‘a major shift in political power and associated ideology in a direction conducive to greater upward mobility for a wider segment of population’ has taken place. Similarly, Acemoglu and Robinson (Reference Acemoglu and Robinson2000) link the extension of voting rights to the spread of educational policies. Despite the establishment of universal male suffrage in 1890, that shift does not seem to have occurred in Spain or was only slowly coming about due to the unequal distribution of wealth and the shortcomings of the political regime. Political and economic elites firmly controlled the Spanish political system by widespread vote buying, coercion and mass fraud, together with promises of individual or collective favours, although these practices weakened over time, especially after the turn of the century (Curto-Grau et al., Reference Curto-Grau, Herranz-Loncán and Solé-Ollé2012; Moreno-Luzón, Reference Moreno-Luzón2007). Lindert's arguments (Reference Lindert1994, Reference Lindert1996) regarding the importance of the degree of political participation in determining the patterns of the provision of public goods are especially relevant here. Powerful elites indeed tend to block the implementation of institutions promoting human capital (Easterly, Reference Easterly2007; Lindert, Reference Lindert2003; Mariscal and Sokoloff, Reference Mariscal, Sokoloff and Haber2000). Although it is true that relevant steps were taken, the Spanish political system, kidnapped by oligarchic interests and restricted political representation, somewhat precluded higher levels of human capital formation. As shown here, the privatisation of common lands from 1860 onwards, which principally benefited the well-off, and their negative effects on the supply and demand for education highlight only another way through which an unbalanced political system hindered economic development.