All roads lead to market integration. Lessons from a spatial analysis of the wheat market in 18th century Spain. Alexandra L. Cermeño and Carlos Santiago-Caballero Working Papers in Economic History 2020-02 ISSN: 2341-2542 Serie disponible en http://hdl.handle.net/10016/19600 Web: http://portal.uc3m.es/portal/page/portal/instituto_figuerola Correo electrónico: [email protected] Creative Commons Reconocimiento- NoComercial- SinObraDerivada 3.0 España (CC BY-NC-ND 3.0 ES) All roads lead to market integration. Lessons from a spatial analysis of the wheat market in 18th century Spain.1 Alexandra L. Cermeño (Lund University)2 Carlos Santiago-Caballero (Universidad Carlos III) 3 ABSTRACT This paper uses newly collected data from a large-scale census (Catastro de la Ensenada) to investigate the scale and causes of market integration in eighteenth century Spain. We use wheat prices observed in more than 5,200 municipalities to analyse the local spatial dependence of prices. We detect several regional clusters in the centre and coasts but find that these were not integrated with each other. We then investigate the first nature, second nature, and demand side determinants of these clusters and find that although geographical constrains like terrain roughness play a negatively significant role, the transportation network allowed connected municipalities to alleviate such obstacles. Our results suggest that unfavourable geographical conditions can be overcome by investments in transportation infrastructures. Keywords: Early modern, market integration, grain markets, prices, geography. 1 Declarations of interest: none. 2 Email: [email protected]. Department of Economic History. Lund University Box 7083. 220 07 Lund, Sweden. 3 Email: [email protected]. Department of Social Sciences. Universidad Carlos III Madrid. Calle Madrid, 126 Getafe. (Madrid) Spain CP: 28903. Funding from the project “Experiencias históricas en la formación de áreas monetarias.” by Fundación Ramón Areces is acknowledged. 1 1. Introduction As states rose, so did markets. The complex relationship between state-building and the development of markets has been analysed in depth by authors like Epstein, who argued that as European countries consolidated and centralised their power, the subsequent reduction of transportation and transaction costs allowed the rise of domestic markets weakening the power of traditional monopolies (Epstein, 2000). Differences in the nature of this relationship in European economies explains in part, according to Epstein, the different paths of growth that they followed and processes like the Little Divergence. The extent to which political centralisation and nationalism translates into economic integration also reflects the ability of political elites to overcome the traditional barriers to the creation of a domestic market (Miladinovic, 2019). However, while the centralization of political power in eighteenth century Spain is well accepted, the existence of a highly integrated market is a more controversial debate in Spanish economic history. Were Spanish absolutist monarchs of the time able to use their political power to create a fully integrated domestic market? Following Epstein’s line of reasoning, the answer to the previous question would have long-lasting effects in the performance of the Spanish economy. There is a consensus in the traditional literature around the idea that the domestic market became integrated by mid-19th century thanks to the transport revolution and the effect of the railroads (Sánchez Albornoz, 1974, 1975, and Peña and Sánchez Albornoz, 1983, 1984).However, more recent studies suggest that the Spanish economy was already relatively well integrated in the mid-eighteenth century (Llopis Agelán and Sotoca López, 2005), and that integration levels during the eighteenth century were in fact higher than those achieved in later periods (Llopis and Agelán and Jeréz Méndez, 2001). Similarly, Reher (2001) argues that economic integration levels in preindustrial Spain were low, but that they increased during the eighteenth century as proved by the reduction in the volatility of prices over time. Grafe (2011) supported the existence of well integrated markets at regional levels, but rejected the existence of a national market in Spain in the eighteenth century. Regarding the sources behind increasing market integration, the most recent literature has highlighted the role of transports, especially the distance to the transport network reducing price differentials between markets (Schulze and Wolf, 2012; Federico and Sharp, 2013) as well as the importance of flows of information (Gao and Lei, 2019). This paper analyses the levels of economic integration in mid-eighteenth century Spain, following the revisionist hypothesis that this was a period of relatively high integration, and assesses the role played by the transportation network. For this purpose, we study wheat prices from the Crown of Castile, at the level of municipalities available from the Catastro de la Ensenada, a territory that represented around 80 percent of current Spain. We analyse the prices of wheat for over 5,200 municipalities and test for their spatial autocorrelation using Local Moran’s Indexes and Getis-Ord statistics which identify patterns on the similarity of data across space. Our results are congruent with the revisionist literature and we find well integrated clusters at the regional level which are not, however, well integrated at the national level. To further investigate the reasons of the regional integration and global dis-integration, we calculate the prices’ coefficients of variation on a random sample of municipalities and those within the 50 km buffer from 2 them. We then test econometrically which geographic and market variables had an impact on this pattern using Ordinary Least Squares and Two Stage Least Squares regressions, and explore the need to adjust for spatial interactions using Spatial Auto Regressive models. We find that these patterns are heavily influenced by geographic variables, but also that the road network developed and used by travellers and tradesmen helped market integration and helped to soften the negative effects of geographical barriers. The paper proceeds by presenting our data sources in Section 2. We explain our methodology in Section 3. Section 4 describes our results and examines drivers of the patterns. We finally conclude in Section 5. 2. Sources and data This paper uses the information from the municipal manuscripts that were created for each municipality in the Cadastre de la Ensenada. The Cadastre was a monumental work carried out during the reign of Fernando VI, whose main purpose was the estimation of wealth for the introduction of a new tax, the única contribución (single tax), to finance the increasing expenditures of the monarchy. The tax had to be paid by each municipality in the Crown of Castile and the estimation of the exact amount to be paid was based on the location’s wealth and a set of objective criteria.4 For that purpose, an army of some 1,000 judges and 90,000 local experts was mobilised to measure every plot of land, count every tree and measure the wealth of every single family liable to pay to the Crown. A significant share of the enormous amount of documentation produced during the process has survived, putting at the disposal of researchers a unique source for the study of eighteenth century Spain. In our case, we consulted the Respuestas Generales or General answers, a questionnaire that the authorities of each municipality had to fill in. We concentrated especially on the questions related to local prices and units of measurement. The most important information that we extracted from the books of the Cadastre was the price of wheat at municipal level. We used the price of wheat to measure spatial economic integration, mainly because of its availability in practically all the municipalities and also for its common use to measure pre industrial market integration in the literature (for example, see Persson, 1999). For the construction of our dataset, we individually consulted the books of 5,271 municipalities, all the locations with information on prices and local units of measurement in the Cadastre.5 Question 14 in the questionnaire inquired about a list of the prices of the main items produced in each municipality and is the one from where we extracted the prices of wheat. However, although the currency was common (reales), there was a high regional diversity of measures determining volume or weight. We used the information contained in Question 9 of the books where the local authorities had to describe the units of measurement that they used in their municipality and from that we standardised all the prices in reales per fanega. Other 4 Although the Basque Regions were part of the Crown, they were not included in the Cadastre as they had a special taxation system. 5 Manuscript books with the general answers can be found in the Respuestas Generales del Catastro de Ensenada and accessed in http://pares.mcu.es/Catastro/ A list and graphic description with all the included municipalities can be found in the appendix. 3 necessary information from the same books and from alternative sources was used in a smaller sample of locations used in our econometric analysis, presented and explained in detail in the specification of the different models in the following section. 3. Methodology We start our analysis by exploring the geographical distribution of wheat