Regional planning insights from a Portuguese bi-regional Input-Output model

In Portugal, the public debate at regional level is typically absorbed in the discussion of asymmetries amongst the ‘interior’ and the ‘coast’. What is often discussed, with political and social relevance, is the extent of the interior’s delay (in terms of development) comparatively to the coastal region, and into what extent the dynamics of the economy, or eventually the ‘bias’ introduced by public policies, contributes to this drawback. Interestingly, however, the Portuguese regional science has miscarried this debate, largely on the grounds that the official statistics do not include this cleavage. In fact, the design of the NUTS II in Portugal splits the country horizontally, forgetting the vertical gap that separates the interior of the coast. Departing from the construction of a bi-regional Input-output model for the Portuguese economy, the main aim is then to assess how the effects of a shock that hits only one of the regions are ‘distributed’ among the two regions. In particular, we intend to analyse at a greater detail the role of the agriculture sector in the Inland Region. Our modelling approach departs from the construction of a bi-regional input-output model for the Portuguese economy, where both the model and the database are formulated in terms of detailed technical parameters, reproducing both inter-sectoral and interregional interdependencies. We consider a rectangular IO model (431 products by 125 industries), decomposing the Portuguese economy into two regions with comparable territorial sizes (the Coastal Region, comprising 44% of the Portuguese surface, and the Inland Region). The Coastal Region is the most industrialized and economically developed one (representing nearly 89% of the Portuguese GDP) and is, simultaneously, the region where most of the population lives (almost 85% of the Portuguese households). In contrast, the Inland Region is more dispersed and founds its economic basis in products such as the ones from agriculture, metal ores, beverages, electricity and transportation services, although the consumption of these products occurs largely abroad. The model is presented at ‘domestic flows’, which means that the use of goods and services produced within the region is distinguished from the use of those produced in other regions/countries. Further, as the model is presented at ‘basic prices’, the VAT and ‘Other Taxes less Subsidies on Production’ are considered outside the core of the IO model. Lastly, the model is ‘closed’ for the private consumption of households below 65 (which is supposed to be endogenous, as it depends on regional employment and therefore on households’ earnings - labor compensations and mixed income - that result from the local production process and simultaneously determines the regional economic variables, including the regional production level), while the older households are considered in the final demand section of the matrix. Therefore, the exogenous final demand is considered to be the one of households whose heads exceed 65 years old plus the one associated with public consumption, investment, (international and inter-regional) exports and changes in inventories. Typical (uni-regional) Input-Output (IO) models describe the inter-sectoral dependencies in a given economy. Multi-regional IO models describe the inter-sectoral dependencies both within the region and between the regions. Thus, in reaction to an external shock, the multi-regional IO framework allows to estimate, e.g., the feedback effects that may be felt in other regions/economies and the ones that then ‘return’ to the region which initially felt the original disturbance. Preliminary overall results illustrate the dependence of the Interior on the Coastal region, and that the (positive or negative) effects of a shock that hits the Interior Region tend to leak to the Coastal Region, while an exogenous event in the Coastal Region tends to see its effects contained within the region. Thus, this analysis can be particularly relevant to policy-makers in dealing with regional and territorial planning, as they are better informed about the root causes of some outcomes. Accordingly, a summary of the key lessons learned and a discussion of their policy relevance, both at regional and national levels will be offered.