Market potential estimates in history: a survey of methods and an application to Spain, 1867-1930

New Economic Geography (NEG) models stress the importance of access to demand as a key driver of the spatial and temporal distribution of economic activity (Krugman, 1991). Therefore, in order to test the theoretical predictions emanating from NEG a sound measure of accessibility is required. In line with Crafts (2005b), this paper constructs market potential estimates for Spain at the province level (NUTS3) between 1867 and 1930 using Harris's (1954) equation. This period is particularly appealing as it was during these years that the Spanish market became integrated thanks to the fall in transport and trade costs. A number of key processes, including the substitution of traditional transport modes and the improvement in transport technologies (the railways and steam navigation), were set in motion. At the same time, the Spanish economy was experiencing the early stages of industrialization in a context characterised by changing trade policies. Interestingly, these long-term processes brought about asymmetric changes in regional accessibility, thus potentially affecting the evolution of Spain's regional economies. The results show that the main changes to regional market accessibility in Spain can be traced to the second half of the 19th century, a period characterised by centrifugal tendencies. The availability of this dataset is essential for undertaking empirical exercises, through the lens of NEG, examining the Spanish experience up to the Spanish Civil War (1936-1939), when an increasing spatial concentration in the country's manufacturing sector and an upswing in regional per capita income inequality took place.