Water and production reallocation in the Spanish agri-food system

Abstract

Multiregional input–output (MRIO) and computable general equilibrium (CGE) models have greatly facilitated approaches to environmental and economic problems in recent years. This paper examines regional reallocation criteria intended to reduce water constraints in the Spanish economy. Our goal is to assess the impact of alternative allocation scenarios for regional production on the country’s agriculture and agri-food industries, and the associated effects on water resources along the whole length of food supply chains, which display significant asymmetries between regions caused by imbalances in the availability of water resources. We design a CGE model using an MRIO database for Spain. Our scenarios are based on increases in the production of water-intensive crops in regions with more abundant water resources and the development of more sustainable food supply chains between farms and the agri-food industry. Our findings point to a series of policy options that could be applied to ensure successful outcomes in both directions.

KEYWORDS:

• MRIO
• CGE models
• water re-allocation criteria
• food supply chain

Acknowledgements

We would like to thank the anonymous reviewers and the editor for their helpful comments.

Disclosure statement

No potential conflict of interest was reported by the authors.

ORCID

Rosa Duarte http://orcid.org/0000-0003-3113-1698

Cristina Sarasa http://orcid.org/0000-0003-4802-600X

Notes

1 The nested production function for the intermediate inputs is not included in the case of foreign regions (the rest of the EU and the rest of the world).

2 Note that just six of them have updated their frameworks to 2010 and even less to other years. Neither is there any updated interregional trade data and sector-specific water data to accommodate an updated MRIO.

3 The term ‘use’ is employed generically here. There is an extensive literature dealing with the concepts and differences of water use types, especially consumptive use and non-consumptive uses, impacts, etc. (e.g. related to the ISO 14046 water footprint definition, in LCA water consumption literature, etc., see ISO_14046, 2014). We follow the approach of the water footprint literature (Hoekstra et al., 2011) and compute both the (direct) blue and green water consumptive use. Blue water is defined as fresh surface and groundwater. Green water is the rainwater stored in the ground and absorbed by crops. Blue and green water have different environmental effects and policy implications, and they can be used together or independently in studies of water and distribution.

4 The method used computes the provincial run-off, which is then aggregated at the level of the Autonomous Communities (i.e. the political regions of Spain). We begin with raster data (1 km by 1 km) on precipitation, potential and real evapotranspiration, and ultimately run-off raster data in Spain (SIMPA, 2010). This we aggregate to the provincial level using ArcGIS. We used supplementary data (FAO, 2019; MAPAMA 2015b, 2015c) to check the total run-off values obtained against alternative measures of water availability.

5 Working with relative indicators in the modelling involves many challenges when being accounted along the supply chains (working with a ratio such as water scarcity ratio, which is again divided by the output to obtain coefficients). An alternative is an approach like Lenzen et al. (2013), which is based on applying initially a pressure ratio (based on consumption per availability) and working with absolute figures of consumption (m³). We also explored water stress indices (e.g. Pfister et al., 2009; Quinteiro et al., 2018) that pose a challenge in modelling since it means working with actual consumption as transformed by so-called characterization factors.

6 For the purposes of this study, the Murcia region has been grouped with the autonomous North-African exclaves of Ceuta and Melilla. Murcia itself accounts for 2.6% of Spain’s GDP, Ceuta for 0.15% and Melilla for 0.14%. In other words, the ‘mainland’ Murcia represents 90% of the ‘grouped’ region. Thus, it is reasonable to abbreviate the label of the discussion results here simply as ‘Murcia’.

7 This amount is less than 10% of total farm output, which was the actual amount of subsidies awarded in Spain in 2005 (MAPAMA, 2005b). We assume the same amount for all Scenarios for reasons of comparability.

8 Higher values for sectoral and regional production are associated with lower physical outputs (production value/price) and price rises. Table SI4 of the SI shows the changes in physical output.

Additional information

Funding

This work was supported by the Spanish Ministry of Science and Innovation (grant numbers ECO2016-74940-P) and the Consolidated group S40_17R of the Government of Aragon (Research Group ‘Growth, Demand and Natural Resources’).