Assessing Macro-economic Effects of Climate Impacts on Energy Demand in EU Sub-national Regions

Abstract

European policy makers are increasingly interested in higher spatial representations of future macro-economic consequences from climate-induced shifts in the energy demand. Indeed, EU sub-national level analyses are currently missing in the literature. In this paper, we conduct a macro-economic assessment of the climate change impacts on energy demand at the EU sub-national level by considering twelve types of energy demand impacts, which refer to three carriers (petroleum, gas, and electricity) and four sectors (agriculture, industry, services, and residential). These impacts have been estimated using climatic data at a high spatial resolution across nine Shared Socioeconomic Pathway (SSP) and Representative Concentration Pathway (RCP) combinations. The impacts feed into a Computable General Equilibrium model, whose regional coverage has been extended to the sub-national NUTS2 and NUTS1 level. Results show that negative macroeconomic effects are not negligible in regions located in Southern Europe mainly driven by increased energy demand for cooling. By 2070, we find negative effects larger than 1% of GDP, especially in SSP5-RCP8.5 and SSP3-RCP4.5 with a maximum of − 7.5% in Cyprus. Regarding regional differences, we identify economic patterns of winners and losers between Northern and Southern Europe. Contrasting scenario combinations, we find that mitigation reduces adverse macro-economic effects for Europe up to a factor of ten in 2070, from 0.4% GDP loss in SSP5-RCP8.5 to 0.04% in SSP2-RCP2.6.

Appendix

1.1 A.1: Regionalizing the GTAP Database

In the following sections, we summarise how the sub-national SAMs have been obtained starting from the GTAP 8 database (Narayanan et al. 2012). Concerning the number of regions in each country, we should keep in mind that the regionalization process is very time-consuming. The process requires to specify all the variables in the original GTAP database at the sub-national level, to balance the sub-national Social Accounting Matrices and to compute the intranational bi-lateral trade flows. Therefore, we adopt a sub-national detail (NUT2 or NUTS1) for the larger economies, such as Germany, France, UK, Italy, and Spain. Small countries such as the Baltic countries, Luxembourg, Slovenia, and Croatia are kept at the national NUTS0 level. Some medium-sized countries like Netherlands, Sweden, Belgium, Poland, and Czech Republic are also regionalised to better represent Eastern Europe, Scandinavia, and Benelux.

It is worth noting that our downscaling method is applied to a global database. Therefore, the database includes information also for 18 regions in the rest of the world; Latin America, USA, Rest of North America, North Africa, Sub Saharan Africa, South Africa, Middle East, India, South Asia, South East Asia, East Asia, China, Japan, Former Soviet Union, Rest of Europe, EFTA, Australia, and New Zealand. For these macro-regions as well, we compute impacts on energy demand in the different RCPs.

1.1.1 A.1.1: Creating and Balancing the Sub-national EU SAMs

The collection of the sub-national information is only a preliminary step to obtain the final database. We use the methodology in Bosello and Standardi (2018) to compute and balance the regionalised SAMs. The methodology is applied in the following steps. In the CGE model, the value added is the sum of primary factors remuneration (labour, capital, land, natural resources). Therefore, the first step of the process consists in disaggregating the value added, originally available at the country level in the GTAP 8 database, to the new regional scale. To do this, first, we match the GTAP sectors with those of our data sources from Eurostat. Then, in each sector the regional shares of labour, capital, land, and natural resources are computed from the sub-national data and used to distribute the respective GTAP data across the sub-national units.

The second step is more challenging as we need to compute intranational trade. This is equivalent to compute the sub-national domestic and imported consumption from the Eurostat information we collected. Indeed, sub-national data on intranational trade is often missing and needs to be reconstructed using different techniques. In our case we rely on the so-called Simple Locations Quotients (SLQs) (Miller and Blair 1985; Bonfiglio and Chelli 2008; Bonfiglio 2008). The formula for the SLQs is the following:

$$SLQ_{i,r} = \frac{{X_{i,r} /X_{r} }}{{X_{i,c} /X_{c} }}$$

(1)

where i is the sector and X the value added, r and c represent the regional and national indexes, respectively. SLQ gives a measure of the regional specialisation in the economic activity. When SLQ is equal to zero, the region needs to import intermediate and final goods from other regions. In the other extreme case, the sectoral value added in the region is equal to the national one and this means that the region tends to export those goods for intermediate or final consumption. Clearly in almost all the cases the SLQ values are in between the two extreme cases. The sub-national shares of domestic and imported demand are obtained by multiplying the national shares times SLQs and then normalising these shares.

The final step consists in the determination of the bilateral trade flows across the sub-national regions. The procedure usually adopted is based on gravitational approaches as in Horridge and Wittwer (2010) and Dixon et al. (2012). By this method, the bilateral intra-country trade flows are estimated using a gravity equation. We also follow a gravitational approach based on the kilometric road distance between each couple of capital cities for the regions within the country. We adjust the trade flows across sub-national regions by using the RAS statistical method (Bacharach 1970) to make them consistent with the aggregate intranational exports and imports obtained through the SLQs.

1.1.2 A.1.2: Splitting the Electricity Sector at the Sub-national Level

In the construction of the SSP-RCP combinations, it is important to represent the electricity sector in a sophisticated manner because the energy sector develops differently according to each scenario, and this has relevant economic implications for the macro-economic assessment. For example, in SSP1 we may expect a strong development of the renewables-based power generation sector and a progressive electrification of the economy while in SSP5 fossil fuels remain important sources for both the electricity sector and the overall economy. Therefore, we have increased the detail of the electricity sector at the sub-national level in the reference year 2007. We use information from the World Electric Power Plants Database (WEPP) (PLATTS 2014) to increase the technological detail in the electricity sector at the NUTS1/2 level. WEPP is a global inventory of electric power generating units managed by S&P Global. It provides information on more than 107,500 plant sites in more than 230 countries and territories and details on plant operators, geographic location, capacity (MW), age, technology, fuels, and boiler, turbine, and generator manufacturers, emissions control equipment, renewable energy units and more. Using the WEPP information, we are able to include in the electricity sector six more technologies at the sub-national EU level: nuclear, fossil power generation, wind, hydropower, solar, and other renewables.

See Tables 5, 6 and Figs. 10, 11, 12, 13.

Fig. 10

Box-plot of climate-induced energy demand trends in EU regions and RCP2.6 (2015–2050% changes left, 2015–2070% changes right)

Fig. 11

Box-plot of climate-induced energy demand trends in EU regions and RCP4.5 (2015–2050% changes left, 2015–2070% changes right)

Fig. 12

Box-plot of climate-induced energy demand trends in EU regions and RCP6.0 (2015–2050% changes left, 2015–2070% changes right)

Fig. 13

Box-plot of climate-induced energy demand trends in EU regions and RCP8.5 (2015–2050% changes left, 2015–2070% changes right)

Table 5 Mapping between EU regions of the ICES model and NUTS 2013 EU code

Table 6 Transmission and distribution (Tr&D) share in the value added of the electricity sector.