The Swedish footprint: A multi-model comparison

Abstract

Sweden has a large per capita carbon footprint, particularly compared to the levels recommended for maintaining a stable climate. Much of that footprint falls outside Sweden's territory; emissions occurring abroad are “embodied” in imported goods consumed in Sweden. In this study we calculate the total amount and geographical hotspots of the Swedish footprint produced by different multi-regional input-output (MRIO) models, and compare these results in order to gain a picture of the present state of knowledge of the Swedish global footprint. We also look for insights for future model development that can be gained from such comparisons. We first compare a time series of the Swedish carbon footprint calculated by the Swedish national statistics agency, Statistics Sweden, using a single-region model, with data from the EXIOBASE, GTAP, OECD, Eora, and WIOD MRIO databases. We then examine the MRIO results to investigate the geographical distribution of four types of Swedish footprint: carbon dioxide, greenhouse gas emissions, water use and materials use. We identify the hotspot countries and regions where environmental pressures linked to Swedish consumption are highest. We also consider why the results may differ between calculation methods and types of environmental pressure. As might be expected, given the complexity and modelling assumptions, the MRIO models and Statistics Sweden data provide different (but similar) results for each footprint. The MRIO models have different strengths that can be used to improve the national calculations. However, constructing and maintaining a new MRIO model would be very demanding for one country. It is also clear that for a single country's calculation, there will be better and more precise data available nationally that would not have priority in the construction of an MRIO model. Thus, combining existing MRIO data with national economic and environmental data seems to be a promising method for integrated footprint analysis. Our findings are relevant not just for Sweden but for other countries seeking to improve national consumption-based accounts. Based on our analysis we offer recommendations to guide future research and policy-making to this end.

Introduction

Current levels and patterns of consumption in developed countries are unsustainable, using too many raw materials and producing too much waste and pollution (Lorek and Vergragt, 2015). This is reflected in developed countries' high carbon, land and material footprints – estimates of the global pressures on ecosystems and natural resources that are linked to a country's consumption. For most developed countries, including Sweden, much of that footprint pressure falls outside of the territory, in the countries supplying Sweden's imported goods (Schmidt et al., 2018 (this issue); Steen-Olsen et al., 2012). This paper compares the global geographic “hotspots” of environmental pressures in Sweden's consumption footprints that are identified in different models. It explores the possible underlying causes of differences between the model results. The research is motivated by the objective to support policy and decision-makers in monitoring Sweden's footprint and provide recommendations for future research to improve the accuracy of national footprint estimates.

Sweden is now one among a number of countries that have produced and analysed their environmental impacts of consumption. The Swedish national statistics agency (Statistics Sweden, or SCB) has published national consumption-based carbon dioxide (CO₂) emissions accounts (carbon footprints) since the end of the 1990s, with current estimates of GHG emissions per product group for 2008–2014 publicly available. In addition, a consistent time series from 1995 to 2009 of data on the CO₂ emissions from Swedish consumption was published by Statistics Sweden in 2015 (Statistics Sweden, 2015) including a comparison of calculation methods using two different models. Earlier pilot studies by Swedish government agencies and research organizations had reported comparable footprint findings (Finnveden et al., 2001; Palm et al., 2006; Naturvårdsverket, 2008).

Work to develop similar consumption-based accounts for numerous countries has also been ongoing over a number of years, examining a wide range of environmental pressures such as the carbon footprint (Hertwich and Peters, 2009; Wiedmann et al., 2010); the water footprint (Hoekstra and Mekonnen, 2012) the land footprint (Weinzettel et al., 2013); and the material footprint (Wiedmann et al., 2015). Footprint results are now publicly available for many countries (Wood et al., 2018).

While the varying results that different models produce for the same footprint indicator may be confusing for communication purposes. There are benefits in examining the outputs of models with varying designs or data sets employed; this variability can be seen as repeated analyses concerned with the same basic set of questions, demonstrating plausibility of a consumption-based accounting approach and raising new policy questions.

This is particularly relevant in Sweden, where a number of national policies and strategies aim to tackle unsustainable consumption. A central component is the Generational Goal, the overarching goal of the national system of environmental objectives. This calls for solving the major environmental problems in Sweden within a generation, without exacerbating health or environmental pressures in the rest of the world (Swedish Environmental Protection Agency, 2012). In addition, Sweden is a signatory to Agenda 2030 and Sustainable Development Goals, with sustainable consumption and production as Goal 12 (United Nations, 2015), and recently launched a national Sustainable Consumption Strategy in December 2016 (Government Offices of Sweden Ministry of Finance, 2016). Regular monitoring of the global impacts of Swedish consumption will be essential to these efforts.