Demand side climate change mitigation actions and SDGs: literature review with systematic evidence search

To strengthen current discourse on acceleration and scale up of the emissions mitigation actions by sector-specific demand side actions, information on the intersection of three dimensions becomes useful. First, what kind of actions help in avoiding, shifting and improving demand for activities/services and resultant emissions to help in deciding choices for actions; second, how these three categories of actions are linked to the wider impact on human wellbeing represented by the Sustainable Development Goals (SDGs) framework; and third, who are the actors associated with these mitigation actions. These three steps become important in the targeted scaling up of actions through policy interventions. This study undertakes a review of the literature between 2015 and 2020 with systematic evidence searching and screening. The literature search has been conducted in Scopus Database. From over 6887 literature in the initial search, 294 relevant literature were finally reviewed to link demand side interventions of avoid-shift-improve (ASI) categories to SDGs. It also maps these actions to actors who can lead the changes. Results show that a wide range of improvement actions are already helping in incremental steps to reduce demand and emissions in various services like mobility, shelter and industrial products. However, ASI categories provide more distinct mitigation actions. All actions need support of innovation, infrastructure development and industrialization. Actions that interact with several SDGs include active mode of transport, passive building design, cleaner cooking, and circular economy. Positive links of these actions to multiple SDGs are overall very strong; however, few trade-offs have been observed. These are mostly related to distributional impact across social groups which highlight the need for policy attention and hard infrastructure design changes. Mitigation and wider benefit outcomes cannot be achieved by individual or household level actions alone. They require the involvement of multiple actors, interconnected actions in sequence as well as in parallel, and support of hard infrastructure. Our results show that in mobility services, policy makers supported by spatial planners and service delivery providers are the major actors. In industry, major actors are policy makers followed by spatial planners and innovators. For buildings, key actors include spatial planners followed by policy makers. Besides these, strategic information sharing to enhance user awareness and education plays an important role in shaping behaviour. Digitalization, information and communication, and interactive technologies will play a significant role in understanding and modifying people’s choices; however, these would also require regulatory attention.


Background
Acceleration in mitigation actions from now through the next two decades simultaneously by all countries and across all sectors is emerging as an imperative from all global assessments (IPBES 2016, UNDP 2016, UNEP 2017, IPCC 2018, EAT-Lancet Commission 2019, HIMAP 2019 to avoid catastrophic consequences and to avoid very high and many uncertain mitigation costs associated with uncertain technologies. One strong argument in favour of wide scale global participation in climate mitigation action comes from the scientific community who are highlighting clear wider developmental co-benefits from climate mitigation (Dubash et al 2013, Bajželj et al 2014, Behnassi et al 2014, Long et al 2016, Ürge-Vorsatz et al 2018, Roy et al 2018a. Development and employment literature (Elasha 2010) see near term national developmental priorities including energy security concerns as urgent priorities and mitigation action as a longterm goal. Sustainable development goals (SDGs) provide a common overarching framing for development that addresses both long-term sustainability and wider human wellbeing oriented developmental goals in the near future for all countries. Climate action is the 13th goal under 17 politically negotiated globally accepted SDGs. Mitigation is one of the climate actions proposed under SDG 13. Synergy and trade-off between mitigation action and the SDGs are now well acknowledged in the literature (Roy et al 2018a, Hoegh-Guldberg et al 2019, Some 2020). An IPCC special report on the 'Global Warming' of 1.5 • C assessed extensive literature to report the indicative linkages between climate action within the SDG framework (Roy et al 2018a) without going into an assessment of net positive and negative impacts. Users of the report expressed its usefulness and a desire for additional assessment of net wider impact. The caption to Table SPM 4 in the IPCC SR1.5 notes the need for further assessment. The IPCC SR1.5 also identifies the lack of adequate studies that address interlinkages between climate mitigation and adaptation and resilient socio-economic transformations.
Our goal in this article is to pick up from the more conclusive findings in the SR1.5, which highlights that demand side mitigation actions are more synergistic with SDGs compared to supply side actions and conduct a deeper investigation using systematic evidence search and screening of the existing literature for a better understanding of the evidence on SDG impacts of demand side mitigation actions grouped under avoid-shift-improve (ASI) categories, and further into actors who can lead the actions.
The ASI categories we borrow from applications in the 1990s in the transport sector. Recently Gota et al (2019) reviewed up to 1500 low-carbon measures in 81 countries, which are grouped under the 'avoid' category as those aiming to decrease the need for transport trips, 'shift' representing actions that enable a shift to substitute modes for service provision, and 'improve' represents actions aiming to look for more efficient appliances. Given the simplicity of the ASI categorisation, recently there is an increasing focus on applying the ASI category beyond the transport sector . Given this background, this study undertakes to prepare a review by a systematic evidence search and screening to address the overarching research question, 'what evidence exists on various mitigation actions implemented in various regions that can be categorized under avoid, shift and improve in demand for products and services, how do they interact with SDGs, and who are the main actors who drive mitigation actions?' The study considers the interaction of mitigation actions with all 16 non-climatic SDGs. Three energy demand sectors (services) chosen are: transport (mobility), industry (variety of products) and building (shelter and commerce).

Method
We have followed a systematic evidence search and screening of literature and evidence. It is performed by using relevant keywords within the scope of the research question to find out all the relevant literature within the Scopus database. Being systematic helps us in removing bias of selecting and omitting literature .
It is very basic, but due to its elegance we organize and segregate the research question in populationintervention-outcome (PIO) elements. The PIO elements of the research question are: Population(s): energy end use sectors (industry, transport and building) mapped into goods and service typologies e.g. industrial manufactured goods such as clothing for thermal comfort , transport for mobility and buildings for shelter for commercial and institutional functions for homes, hospitals, schools and workspace. However, we use sectors and services interchangeably in the article.
Intervention(s): a range of mitigation actions relating to various sectors mentioned in the population above. These mitigation actions are categorized as changes at technical, behavioural and infrastructural and/or systemic level.
Outcome(s): synergies and trade-offs with all 16 SDGs with goal 13 (climate action) as climate mitigation actions in energy demand sectors are the entry point of this study.

Literature search
To identify demand side mitigation actions for enduse sectors such as transport, industry and building, search queries (annex A) are developed on the sector-specific key elements PIO (population (Query 1), intervention (Query 2) and outcome (Query 3)) of the research question. We divide outcome-related search terms into two parts: Query 3a and Query 3b; the first one focuses on various demand side actions for mitigation and the second one is to constrain results to those relevant to SDG linkages. The search was performed in Scopus. We are aware that these same search strings would have fetched different results in other databases (Bramer et al 2017) but we have used only Scopus because it is more user-friendly in terms of downloading search results directly in .csv format than other databases like Web of Science (Core Collection). We have limited the study period between 2015 and 2020 to align with the SDGs and for the ongoing IPCC Sixth Assessment Cycle (AR6). Annex A provides a list of search terms used for the three sectors.

Screening
Articles from the above search were considered for inclusion in three successive levels following selected inclusion criteria (annex B). The first step was title screening followed by abstract screening and finally full text review. In case of uncertainty during the title or abstract screening, the article was included by default in the next step of screening. After the abstract screening, 12 articles randomly selected from each sector were checked by all the four authors for consistency and agreement between the authors and were recorded using the Kappa test. The Kappa values are ranged in between 0.63 and 0.71 (k = 0.71 for transport: SS vs JR; k = 0.63 for industry: ND vs JR; k = 0.66 for building: MP vs JR). 5 The scores indicate moderate agreement.
The final step was a full text review of short-listed articles for inclusion in the final analysis. The articles were distributed by sector among two reviewers: transport-JR and SS, industry-JR and ND and building-JR and MP. General inclusion and exclusion criteria were set a priori for all the three sectors at the different stages of screening (annex B).

Data/information extraction
Articles included for final analysis are thoroughly reviewed. We used a detailed coding sheet (see the supplementary file which is available online at stacks.iop.org/ERL/16/043003/mmedia) to record the mitigation actions, the country context, the SDG linkages and the actors along with all other necessary information related to the articles like the title of the article, publication year, source and the DOI.

Synthesizing data/information
We grouped various demand side actions from the supplementary file into the ASI category following Creutzig et al (2018). The ASI category for all three end-use sectors in this paper is grouped as follows. The Avoid catergory includes actions aimed at avoiding the demand for high-emission intensive services (mobility, manufactured product, shelter and building services including cooking); actions in the shift category help in substituting demand for high-emission intensive services with low/no intensive ones; and improve category actions are aimed at improving the energy/emission intensity of a service type. Two reviewers from each sector categorized the interventions into the ASI category and in case of ambiguity, a third reviewer was consulted to solve the discrepancies. We then mapped the SDGs' impacts of individual ASI interventions (see supplementary file). SDGs are interlinked, so trying to achieve one goal (here SDG 13) have wider impacts (synergies and trade-offs) on other goals (other 16 SDGs, except SDG 13). Therefore, it is important from a policy perspective to understand these wider impacts that the demand side mitigation actions imply. We have mapped the wider impacts from the implementation of demand side mitigation actions (SDG 13) against the other SDGs, for each of the three end-use sectors (see table 1, figures 3-5 and supplementary  file). Then, we tried to see which actors can be associated with these mitigation actions. This becomes important in targeted scaling up of actions through policy interventions. These findings are presented in section 3.
While reviewing the articles, we found interesting inter-sectoral linkages that can aid in taking the mitigation actions. We report the findings of the inter-sectoral linkages and digitalization in section 4. Figure 1 presents the workflow chart.

Descriptive statistics
The search queries fetched a total of 6887 studies. Annex C provides details of the articles included after each screening stage. Below we describe the screening stages for each sector. A total of 294 articles are included for detailed analysis. Studies included in the review spanned a wide geographic scope representing over 67 countries with a large number of studies from Europe, Asia and North America (figure 2). Considering all the sectors, most studied countries are China (n = 25) followed by USA (n = 21), UK (n = 17) and India (n = 15). In total, 18 studies had a wider scope focusing either at the global level, on a specific region or on developing countries in general. For the transport sector, most studied countries are China (n = 10), followed by USA (n = 9) and India (n = 8). Out of 107 transport sector studies included in the review, 50 are developed country studies while 47 are developing country studies. For the industry sector, the most studied countries are China (n = 6), followed by India and Italy (n = 2). Out of 59 studies for industry included in the review, 7 are developed country studies while 11 are developing country studies. For building sector, most studied countries are USA (n = 12), followed by China (n = 9). Of the 128 studies included in the review or building sector, 95 are developed country studies while only 19 are developing country studies.

Transport
In this sector, we have grouped mitigation actions in service delivery into the ASI category (see annex D) following Creutzig et al (2018). The Avoid category includes actions aimed at avoiding the demand for mobility services: compact urban planning which helps in avoiding mobility and reducing vehicle ownership. Shift category actions help in substituting demand for mobility services with a low/no emission intensive one: increasing the use of active travel and more use of public transport. While, actions in the Improve category are aimed at improving the energy/emission intensity of a mobility service type: shared mobility, use of electric vehicles (EVs)/zero emission vehicles, hydrogen buses, etc.

ASI-SDG link
Mitigation actions in the mobility service sector deliver multiple benefits and therefore show synergies with several SDGs (table 1). Compact urban planning and reducing vehicle ownership cater to sustainable cities and reduce air pollution (SDG 11). Active mobility requires the development and strengthening of partnerships (SDG 17) for planning (Macmillan et al 2020) the infrastructure. Shifting to active modes such as walking, cycling, etc, reduces mortality and provides health benefits (SDG 3) but has a collision risk if not supported by separate lanes (Doorley et al 2015). Using active modes (cycling) generates income for the local communities (SDG 1 and 8) and for commuters, these are very low-cost methods for accessing basic services (Macmillan et al 2020) for developing countries. Active modes can reduce gender inequities in access to basic services, healthcare and education (SDG 5). Education and awareness programs aid in understanding the environmental benefits of shifting to public transport and e-vehicles (Bigerna et al 2019). Urban environments that enable active travel modes for trips have the potential to reduce physical and financial barriers to participating in education for women thereby catering to SDG 5 (Macmillan et al 2020). A study in Canada (Mitra and Nash 2019) has shown that providing a bike lane facility improves the chances of females (SDG 5 and 10) commuting by bike. The use of public transit and active modes demands proper infrastructure (SDG 9) and saves energy (SDG 7). Also, Sjöman et al (2020) mentioned that transit-oriented infrastructure (SDG 9) helps in avoiding ownership of private vehicles. However, in some developing countries weather conditions and unreliable connectivity affect the lack of incentives to improve existent infrastructure related to public transportation (Bonasif 2017) and active modes. Gilderbloom et al (2016) pointed out that shifting to active modes and reducing car demand preserves land (SDG 15) that would have been otherwise used to construct and maintain parking garages and surface parking lots.
Improving service efficiency (e.g. car sharing) or fuel switch (e.g. hydrogen buses; EVs)  helps improving mobility service demand with relatively less emission. Sharing mobility services has come a long way now but commuters with a private interest in driving their own car need to understand the importance of sustainable consumption (SDG 12) and act altruistically (Mehdizadeh et al 2019). Using low or no-carbon fuel and efficient car help in reducing air pollution which provides direct health benefits (SDG 3) (Yang et al 2018). Also, studies have shown that digitalization has helped women to use more carpooling due to the built-in safety features (SDG 5, 10 and 11). However, subsidizing EVs for greater adoption may lead to higher sales among active travelers (Rudolph 2016. Out of 107 selected studies in the final analysis, almost all studies point toward benefits across the various dimensions of sustainable development from mitigation actions for transport. Overall, no direct linkage has been found with SDGs 2, 6 and 14. Most of the linkages are with SDGs 9 and 11, clearly showing the relevance of innovation and infrastructure for sustainable transportation, followed by SDGs 16 and 17, clearly depicting the crucial role of actors like government and planners in ASI. It is apparent from figure 3 that actions in the shift category and a few in the improve category (shared mobility and EVs) have more evidence for the SDGs link.

Actors and examples of mitigation actions
Spatial planners and policymakers are identified as the major actors for helping in avoiding the demand for mobility services. There are various examples where imposing high congestion charges for driving in the central parts of the city like in London (Ahmad and Puppim de Oliveira 2016), rewarding the (voluntary) forfeit of a (second) car in the household (Schoenau and Müller 2017), and imposing high parking fees and high taxes on cars (Ahmad and Puppim de Oliveira 2016) by local governing bodies may have helped reduce ownership of a vehicle. Imposing high parking fees is supported by both developed (Langlois et al 2015) and developing (Becker and Carmi 2019) countries. Spatial planners (e.g. urban planners and designers) (Stojanovski 2019) can adopt long-term strategies like increasing housing supply near the workplace (Schneider and Willman 2019) or transit stops (Langlois et al 2015) to avoid mobility demand. The Seoul Smart Work Center is an initiative for all government employees to work closer to their homes. In 2015, 30% of the employees were covered under this scheme (Shmelev and Shmeleva 2018).
Indian mega cities such as Mumbai and Kolkata have historical advantages of a well spread-out infrastructure of public transport. Kolkata, Pune and Delhi's comprehensive mobility plans and master plans have set a target of achieving 90% modal share in the form of public transport (Roy et al 2018b). The involvement of actors at multiple levels is necessary to shift mobility demand from private mode to public modes. For example, spatial planners (e.g. road planners) or policymakers can adopt various actions like reduced road space for cars; introduce programs like 'car free Sundays' in the city center as done in Bristol; and communicate the environmental benefits of using sustainable transport (Mir et al 2016), while service delivery agents can improve the quality and frequency of service (Trinh and Linh 2018). Developed country studies suggest innovative design for bus services considering user needs at the bus stop (Hildén et al 2016, Mozos-Blanco et al 2018 while developing country studies point out that public transport may not a preferable option due to poor schedule, dirty stations (Lelono et al 2018) and also poor-quality service like rude staff, low-quality buses, crowding (Trinh and Linh 2018), less safety and careless bus drivers. Therefore, innovation and service delivery through service delivery agents have different roles in developed (innovative service design, digitalized mobility (Sjöman et al 2020)) and developing countries (professional training to bus drivers and ticket collectors, providing bus route information, timetable (Trinh and Linh 2018)) to boost demand for public transport. Shifts toward the public transport system in India are backed by policy and practice, as well as an investment model through well-defined private sector participation along with the introduction of joint strategies for energy efficiency of equipment and a switch to electrification (Roy et al 2018b). However, in developing countries like Ghana and India, people with higher education use less active transport, as owning a car indicates higher social status (Acheampong and Siiba 2018); therefore, individuals' own lifestyle choices are also important in making mobility service decisions.
The roles of universities, educational institutes and companies are important both in raising awareness and helping to changing the behavior of individuals. For example, companies can provide monthly prepaid public transport passes rather than private transport allowances (Ahmad and Puppim de Oliveira 2016) as done in Seoul and Tokyo which encourage staff to commute by public transport. To promote active travel, policymakers and local stakeholders should advocate the environmental benefits of walking and cycling/biking (Ho et al 2017, Schneider andWillman 2019). Policymakers and spatial planners like road planners should provide better infrastructure and provide a user-friendly environment to encourage people to walk and bike for their non-work travel needs (Ramezani et al 2018, Carroll et al 2019, Useche et al 2019. While actors like companies (corporate offices) can motivate their employees to use active travel like in Boston, USA, 14 companies are motivating employees to commute to work using bikes (Wunsch et al 2016). There are several case-studies (Aittasalo et al 2017, Fenton 2017 on promoting active modes available for developed countries. Already in many developed countries like Canada and the USA, actions from policymakers and public investment in infrastructure have helped in upscaling active modes (Ramezani et al 2018, Mitra andNash 2019). Interesting to note here that there are very limited studies (figure 3) that discuss promoting active travel in developing countries. These studies (Wethyavivorn and Sukwattanakorn 2019) suggest scaling up active modes for solving last mile problems.
Multiple actors have roles to play in improving mobility service demand. In most of the metropolitan cities in both developed and developing countries, Uber/Ola/Lyft/similar service providers are extremely common. These actors have introduced a vehicle pool service facility where commuters traveling in the same direction can share rides and save money. Policymakers can also promote shared rides through policies (Ceccato and Diana 2018). Sharing mobility services for children's school travel is the responsibility of the parents as household actors (Mehdizadeh et al 2019). In Indian cities like Kolkata, school-bus and carpools are the most common means for transportation of school children but the parents of school children have shown concern about poor service quality, safety, staff behavior, journey time and waiting time (Prasad and Maitra 2019). Most schools prefer a school bus to car pools due to safety issues (South Point School 2019). Actors like employers are providing car sharing platforms for their staff. Therefore, service delivery agents need to look into safety issues to scale up shared ride among school students. Policymakers (e.g. national or local government) support in terms of financial aid (subsidies), infrastructure (charging stations, proper roads) and training facilities for the drivers are very crucial in upscaling EVs , Ahmed and Karmaker 2019, Tu and Yang 2019 in both developed and developing countries. In countries like South Korea and Chile, the government provides subsidies for the adoption of electric taxis (Aymeric andFrançois 2017, Kim et al 2017). However, to date low penetration of EVs has been a problem, but the reality is changing very fast. The reasons for low penetration are high battery cost and low availability of charging points. In Chile, as of 2014, only 136 vehicles (0.003% of the total fleet) are electricity-fuelled (Aymeric and François 2017). It is also important for innovation and service delivery providers to provide adequate charging points to support the penetration of EVs. In Brazil, the Itaipu hydroelectric power plant has established an electric-car sharing platform for its employees (Vanzella et al 2018). Households/individuals can play their role by paying a higher price for low polluting travel. A study (Bigerna and Polinori 2015) designed to find the willingness to pay for hydrogen buses in Perugia, Italy reveals that commuters are willing to pay 30%-60% more than the single-trip bus fare.

Industry
Demand reduction for industrial products can happen at two stages, intermediate demand for the products used as input in other manufacturing processes and final demand from the end-user. However, it is identified to happen mostly outside the industry sector (Fischedick et al 2014). We identified 14 mitigation actions following a full-text screening of 39 articles. These actions for the industry sector are presented in annex E. These actions are broadly grouped into three categories, avoid demand/consumption of virgin materials and inputs by using more recycled and by-products, reduced emission intensity by shifting to the low-carbon or renewable energy sources and modern manufacturing process and improved material efficiency and energy efficiency. The most commonly studied intervention is clustering of the industries through an eco-industrial park (EIP) and industrial symbiosis (IS) (Jin et al 2017, Guo et al 2018, Fraccascia 2019. In IS when a group of firms located in the same area and interconnected to share water, energy and by-products, the operating efficiency of individual firms improves (Bellantuono et al 2017). The symbiotic relationship among firms in EIP and IS avoids material use and facilitates the use of by-products and waste recycling (ElMassah 2018), thereby reducing the consumption of materials and energy and consequent reduction of GHG emissions. Two important demand side mitigation actions in industry are reducing emission intensity and improving energy efficiency at the production process level.

ASI-SDG link
Based on the literature review the demand side mitigation actions for the industry sector are mapped against eight SDGs. From the selected literature goal numbers 1, 2, 4, 5, 10, 14, 15, 16 could not be linked directly. Out of 59 selected studies for final analysis only 12 studies have established direct linkages between SDGs and demand side mitigation actions for the industry. The remaining studies, while do not mention SDGs specifically, point toward benefits across the various dimensions of sustainable development. Goal number 9 directly addresses sustainable industrialization by fostering innovation. In our selected studies almost all the 14 actions are contributing to achieving this goal (Bertarelli and Lodi 2019) mainly through technological innovation (Pigosso et al 2018) and improved energy and emission efficiency (Wang et al 2018b). Sustainable production (SDG 12) with efficient resource management, an important impact of demand side actions with material efficiency (Abreu et al 2017, use of by-products in a symbiotic production system (Bellantuono et al 2017) and circular economy . We have been able to link a total of 8 actions with SDG 12. Cooperation and communication among firms and financial stability with a suitable business model seem to be the most important aspects of clustering of the industries (Menato et al 2017). This also helps attain SDG 17 which addresses partnership to achieve sustainable development through domestic cooperation (see table 1, figure 4 and supplementary file).
It is apparent from figure 4 that avoid category actions have the strongest connection with the SDGs. If we analyze these interventions in the context of developed and developing countries, we can see that avoid category actions like IS and EIP are important demand side mitigation actions for both developed and developing countries. However, the importance of avoid (circular economy), shift (use of low carbon energy sources) and improve (technological efficiency) has a larger importance in developing countries. Smart energy management and smart energy systems have less relevance in developing countries (figure 4).

Actors and examples of mitigation actions
Industrial symbiosis and EIP has emerged as one of the important action from the demand side both in developed and developing countries. However, the driving force of IS is cooperation among industries operating in IS/EIP, information sharing (ElMassah 2018) and shared support services (Fraccascia 2019) which are influenced by various technical, economic, and legal factors . The heterogeneity of the existing firms in a cluster is key for a successful aggregation (Bellantuono et al 2017). At the organizational level the roles of spatial planners such as operators of the firms, IS facilitators, and policymakers become crucial to facilitate a convenient sharing platform for the implementation of a stable IS (Bellantuono et al 2017, Fraccascia 2019. In order to extend the possible use of by-products and for better energy management, linkages with research organizations are important (Guo et al 2018). Among these studies, two studies have argued that tax deduction instead of subsidy can be a more effective instrument to encourage the use of energy-efficient and clean technology (Pillay and Buys 2016, Bertarelli and Lodi 2019), which has also been supported by theoretical models for a long time (Pearce and Turner 1990). In this case the role of a regulatory mechanism to strengthen the market instruments and optimum policy defining the role of policymakers such as government is crucial.
At the user end, actors involved in innovation and service delivery facilitate through digitalization to create an alternative form of provisioning system, say e.g. for clothes swapping through various on-line applications (apps) and also creating pop-up marketplaces (Holmes 2018). Dedicated websites for clothes swapping like 'swishing.co.uk' , 'swopped.co.uk' or apps like 'This for That' have provided new marketplaces for used items of clothing (Holmes 2018) and also contribute to avoiding the demand for new clothing and hence textile industry products. The use of digital media in daily activities like switching to an electronic newspaper is reducing demand for paper newspapers (Permatasari et al 2018).

Buildings
While the ASI category has been extensively applied to the transport sector, very few studies have discussed this for the building sector . Demand reduction from the building sector includes actions that avoid the need for floor area/building footprint or those which reduce the energy consumption for space heating/cooling, illumination from lighting, cooking and use of various energy using appliances. Actions that reduce the need for energy intensive materials for building construction such as cement could include substitution or use of alternate materials such as fly ash, wood, bamboo or other local materials. For the purpose of this study, these actions are excluded. We focus only on the emissions from building operations or services. The avoid category refers to actions that reduce the need for heating or cooling such as passive construction including the use of daylighting, shading, building form and orientation and use of natural ventilation (annex F). Given that many of the green building mitigation actions such as passive design, cool roofs, and green roofs avoid the demand for artificial lighting, heating or cooling services, we include green buildings in the avoid category.
The shift category includes actions such as switching to more alternative substitutes, cleaner cooking fuels and technologies and district heating or cooling. Shift to cleaner cooking fuels is a significant shift action contributing to substantial energy efficiency improvements and SDG benefits discussed in the subsequent section. Purchase or replacement of efficient appliances that fall into the improve category (annex F) are shown to deliver substantial energy savings; however, some studies in Japan, China and Taiwan reported rebound effects undermining total potential energy savings (Mizobuchi and Takeuchi 2016, Su 2019). Environmentally sustainable design including the use of daylight, management of circulation area could significantly reduce the need for heating/cooling and artificial lighting (Lourenço et al 2019). Several studies reported the use of information and communication technology (ICT), Internet of things (IoT) devices that provide real-time consumption data provided useful information to households stimulating energy saving behavior. Such automated systems achieve significantly higher results when matched with conscious and sustainable consumers (Fabi et al 2017). Behavioral aspects could have significant impacts on energy consumption and need to be accounted for (Walzberg et al 2019), which makes it necessary for decarbonizing strategies to go beyond a technology centric approach and include social dimensions (Sodagar and Starkey 2016).
Studies on passive building design, energy saving behaviour and efficient appliances were found for both developed and developing countries however, not surprisingly, there were more studies for developed countries compared to developing countries. Studies on heat pumps, ICT and renewable energy/distributed generation were found mostly for developed countries indicating a higher level of technology penetration.

ASI-SDG links
Demand side actions for buildings deliver multiple benefits and therefore show synergies with several SDGs (table 1). However, there are stronger synergies with SDGs 3, 7, 9, 11 and 12. The highest synergies are observed with SDGs 7 and 9 (figure 5). Sustainable buildings through their numerous mitigation actions such as passive design measures, green and cool roofs, greening systems, water efficiency and other actions enhancing thermal comfort, and financial benefits have the opportunity to deliver multiple positive impacts especially enhancing the health and wellbeing of occupants and therefore we see positive interaction with SDG 3 (Balaban and de Oliveira 2017, Alawneh et al 2018). Sustainably built infrastructure is central to sustainable communities and cities (SDG 11) and buildings, especially large institutional buildings contribute to resilient communities and cities. Studies highlight the need to interpret these benefits with caution as in some cases, certified green buildings or green products could lead to more adverse indoor air quality and health hazards (Steinemann et al 2017). Some studies also show that the performance of green buildings could decline over time undermining the initial benefits (Zaid et al 2017) highlighting the need to undertake long term green building performance evaluations.
Building services are essential for wellbeing and therefore these have a strong impact on equity. Elements include thermal comfort, adequate lighting, heating and cooling and other services. The use of informal biomass for cooking remains a persistent issue globally despite advances in the energy sector and therefore enhanced access to clean cooking fuels and stoves can contribute significantly to overall wellbeing (Poblete-Cazenave and Pachauri 2018). Such a transition could contribute to achieving SDG 7 on universal clean energy access, but also improve gender equality (SDG 6), and health outcomes (SDG 3) (Rosenthal et al 2018, Batchelor et al 2019. Improved illumination and indoor air quality in schools can improve student's learning outcomes and provide opportunities for energy savings (Hu 2017, Lourenço et al 2019, for low income housing, this could also improve women and child health (Rosenthal et al 2018). Similarly, meeting the demand with renewable energy resources caters to SDG 7 on energy security. Technologyled or behavior-led mitigation actions that promote production and use of efficient devices further targets SDG 12 (Withanage et al 2016). Research in energy systems, smart-grids, smart-meters and other technology combinations spur green innovations contributing positively to SDG 9. While efficient appliances have been discussed in most studies, few papers discuss the adequate levels of appliance ownership to ensure decent living standards for societies. A large share of the population in developing countries lacks access to air conditioning and cooling devices. This could further exacerbate energy poverty adversely impacting SDGs 3 and 7 (Mastrucci et al 2019).

Actors and examples of mitigation actions
A number of actors influence the building sector ranging from policymakers (government and quasigovernment agencies) to a number of spatial planners (private stakeholders like real estate developers, builders, architects, designers), local stakeholders (like regulatory agencies) and households. Figure 5 shows the various building related mitigation actions and the actors that can influence change. Policymakers at all levels can influence and support the adoption and upscaling of building interventions. At the national and state level, these include building codes and policies, appliance standards and efficiency targets. National building codes have been successful in most implementing countries. At the local level, planning authorities, independent architects and building professionals exert significant influence on local regulations and targets including building design guidelines and overall urban built form. Accelerated adoption of green building standards also depends on an enabling market environment (Teng et al 2019). Rating agencies or government departments could facilitate green buildings or related sustainable mitigation actions by setting effective standards (Darko et al 2017), through supportive policies technical support and financial assistance (Balaban and de Oliveira 2017), by working with service delivery actors like industry stakeholders to remove regulatory bottlenecks. The private sector also exerts significant influence as a supplier of building services. This includes spatial planners and innovation and service delivery actors who are diverse, ranging from large design and construction companies, architects, designers, building professionals, appliance companies and the emerging players in the IoT and smart metering market.
As the final segment in the chain, households are key agents of change in the energy service demand space (figure 5 and annex F). Actions taken by households could include reduced wastage (switching off lights and appliances when not in use), shift to more alternative modes of service provision through feedback from installing smart devices and adopting renewable energy. Behaviour is influenced by diverse factors including conformity to others (Walzberg et  Studies show participants lack awareness or have inadequate or incorrect information regarding appliance use leading to wasteful behavior (Withanage et al 2016). Cheah et al (2018) propose a framework based on three criteria: (a) reframing sustainability as a future of others' to 'present for us' , (b) responsible consumption and (c) consumption feedback to encourage responsible electricity consumption. Providing online and active feedback on energy consumption can drive behavior change (Ahvenniemi and Häkkinen 2019). These can be overcome by actions that can direct change through 'product or system led innovations' or 'behavior-led' to direct users toward more sustainable behavior. The diverse stakeholders in the building sector often make upscaling challenging as many of these solutions require cooperation among multiple actors. Often a single organization/actor can make a powerful change as evidenced through the example of successful actions in educational institutions. Sustainable decarbonization of the building sector hinges on a combination of technology and social measure and would require ambitious and simultaneous effort from policymakers, spatial planners and households to facilitate an equitable transition (figure 5).

Discussion
We could successfully bin the mitigation actions in ASI categories. How each one is linked to SDGs is also done. We could link each of these to SDGs and identify actors for each mitigation action for each study. Binning of the literature by country was also carried out. Table 1 provides concrete examples in each ASI category. However, in many policy dialogues, knowledge dissemination workshops we came across two most frequently asked questions (FAQs): which action is connected with the maximum number of SDGs and how to compare that across actions? And the second question is that the larger the spread of the actor network, the harder it is to coordinate, so how can we know the size of the actor pool that drives various actions? So, in this section, we present three graphical views to answer these FAQs for three sectors (section 4.1). Also, we discuss inter-sectoral linkages (section 4.2) and the impact of digitalization (section 4.3). Figure 6 shows that all of the shift category actions are positively linked with more than 50% of the SDGs; shared mobility and EVs in improve categories are positively linked to 50% of SDGs but are also not devoid of negative linkages. All actions in avoid categories are linked to less than 50% of SDGs, but are devoid of any negative linkage and also need a lower spread of actor network and so are easier to implement with relatively less coordination need. Public transport requires the involvement of 100% of the actors and therefore needs more effort in coordination.

Sector-specific discussion: relative SDG link and spread of actor network
In figure 7 we notice the least trade-offs with SDGs of various actions except for the two in the improve category, materials efficiency and smart energy management. Avoid and shift categories of actions have all positive SDG links and no tradeoffs but with a maximum of 50% of SDGs. Except for technological efficiency improvements, actor network needs are also relatively smaller and less than 50%.
In figure 8 it is interesting to note that in the building sector interventions, the avoid category have the maximum of positive SDG links with some minor trade-offs, but the actor network is also relatively larger compared to others. However, in the improve      category, the actor network needed for efficient appliances is more compared to others in the category, but SDG links are also more. Some actions such as clean cooking, heat pumps and district heating have all positive SDG links. This analysis provides very useful insights which are novel in this study and provides a useful guide for policymakers, and for that matter for any of the decision makers.

Interaction across sectors
For actions toward sustainable urban mobility, adequate and well-designed road infrastructure and compact urban planning are essential. especially public transport and non-motorized transport. This helps in attracting new residents because of reduced transportation costs, and increases their willingness to pay for the property. This increases housing property prices and governments can earn additional tax  revenue (Shen et al 2018). A case-study from Seattle based on four completed bus transit-oriented development projects reveals that the value of the houses within 0.5 miles radius of the projects has increased significantly (Shen et al 2018). Also, increasing travelers' willingness to adopt active modes largely depends upon the built-in facilities (walking and biking lanes, the distance between buildings and transit shops). Both transit-oriented development and acceptance of active modes demand joint action from the construction industry and the planners (road and urban). Also, the construction industry is innovating ways to build infrastructure that can facilitate the needs of sustainable mobility like the use of rubberized asphalt concrete to improve the performance of pavements (Wang et al 2018a) and the use of recycled concrete aggregate as the base course of new pavements (Reza et al 2018).
The industry sector has obvious backward linkages in supplying end-use products to the transport and buildings. Consequently, demand changes in these sectors directly impact industrial production. For example, circular economy principles such as recycling of construction waste, sharing economy such as shared workspaces, etc reduce the demand for built space (Liu and Lin 2016). The transition toward clean mobility impacts the demand for non-ferrous metals, in some cases leading to an adverse impact on material consumption and in a broader context of environmental sustainability. EV will increase demand for lithium, cobalt and nickel for battery manufacturing (Meskers et al 2019) and manufacturing of lightweight vehicles will result in increased demand for aluminum (Soo et al 2018). Industry being part of the societal transition plays an important role in enabling this transformation. An important linkage between buildings and industry is the use of waste heat from industrial production for heating buildings (Safaei et al 2015).
A number of studies have examined the life cycle impacts of building materials and implications for the construction industry (Muller et al 2019). Sustainable construction practices including the use of alternate building material options include substitution with fly ash, wood, and local materials, recycling and re-use of construction waste and extending building lifetimes have a significant impact on the industry sector through the demand and production of energy intensive materials such as steel and concrete (Dhar et al 2020). Urban planning and design actions including compact urban form, transit-oriented development, and vehicle to grid are examples where buildings and transportation intersect.

Digitalization
Information and communication technologies have emerged as one of the biggest enablers of various demand side changes in all the three sectors through improved interconnectedness, information sharing and use of artificial intelligence. Shared mobility is used by commuters is gaining wider acceptance and it is made possible due to digitalization. Various 'apps' are available for sharing cabs (e.g. Uber, Ola, Grab, Lyft), owned cars (e.g. sRide in India, Bahon, Pathao in Bangladesh, BlaBlaCar in Europe) and bikes (e.g. Mobike in China, Gojek in Indonesia, Thailand). The biggest carpool and bike pool app in India, sRide has made office commutes convenient and led to 80% of the transportation cost savings for commuters (sRide 2019). This app offers various features like a safety toolkit, real time tracking and 24 × 7 customer support. They help coordinate transport for commuters on the same route. Commuters from densely populated Indian cities like Mumbai commented that they find their service on sRide (2019) hassle-free, timely, economical and safe. Liu (2018) reported that commuters in China used Mobike to travel a total distance of 5.6 billion km in 2016 nearly equivalent to 1.26 million tons of carbon emissions. Smart transportation with the help of IoT is not only making daily commute economical, easy and safe but also reducing carbon footprint (Piramuthu and Zhou 2016).
For industrial demand, digitalization itself is not a mitigation action, but it enables optimization of energy use in industrial processes through enhanced energy efficiency with improved information sharing (ElMassah 2018), help in the development of additive manufacturing (Nascimento et al 2019) and industrial communication networking system (Liu et al 2015). At the consumer end, digitalization has enabled the creation of an alternative form of provisioning for clothes swapping through various apps, and also creating pop-up marketplaces (Holmes 2018). Dedicated websites for clothes swapping, like 'swishing.co.uk' , 'swopped.co.uk' or apps like 'This for That' , have provided a new marketplace for used clothes (Holmes 2018) and also contributed to avoiding the demand for new clothes and hence the textile industry. The use of digital media in daily activities like switching to electronic newspapers is reducing demand for paper newspapers (Permatasari et al 2018).
The building sector has seen an increasing focus on 'smart buildings' , 'intelligent buildings' that include a significant level of automation at the enduser level, device-level or system level. This includes the use of ICTs that can measure, record, and influence operations (Fabi et al 2017). A number of studies show smart meters and sensors that provide real-time information on energy consumption or enable digital building-occupant interactions to enhance awareness of users and can assist the shift toward sustainable consumption (Jensen et al 2018). A recent study also showed that state-of-the-art equipment that provides specific feedback on user behavior can enable sustained behavior change.

Conclusions
For a long time, the discourse on climate change mitigation action has focused on the supply side mostly around technological advancement and/or fuel change. With rising urgency in accelerated mitigation action in a short period of time, literature is also focusing on the role of demand side solutions in climate change mitigation. Specifically, in the transition phase how demand can be avoided, shifted or improved to scale up and accelerate the momentum of mitigation actions. In the forthcoming IPCC Sixth Assessment report, an entirely new chapter is dedicated to demand and services besides usual end use demand sectors. The report also frames climate change mitigation in the context of sustainable development. The systematic search for evidence in the present study is based on systematic screening of literature relevant to the demand side of the mitigation actions for transport, industry and building and makes a unique contribution by synthesizing the existing studies from 2015 by three layered analytical framing. This study generates information that can help the new assessment report.
A compilation of demand side actions across sectors needs a common and comparable approach. The ASI category is one such framing that offers the advantage of comparability. While it has so far been applied extensively to the transport sector, the paper attempts to use this framing to collate actions in all three major demand side sectors-transport, buildings and industry. Secondly, the aim was to understand how various actors need to play their roles in various categories and how such actions are linked to wider goals under SDGs.
Findings clearly show that upscaling actions for mitigation would require combined efforts among policymakers, and various stakeholders like households, individuals, city planners, and educators. There is a clear scope for exploring new ways of organizing businesses/services taking advantage of options that create small local supply chains and also with a role for digitalization megatrends in certain fields of actions. Also, cross-sectoral collaboration, resource sharing and information exchange accelerate mitigation outcomes. The study identifies the key enabling factors and actors that can facilitate such a change. The role of governments in incentivizing these solutions through targets, performance standards, and regulatory instruments like tax exemption, subsidy, new pricing policies, monitoring compliance is obvious. A survey of 43 professionals on green buildings (Darko et al 2017) indicated that implementation of standards for design and construction was the most significant factor driving the adoption of green building technologies, followed by factors such as benefits of energy efficiency, health and wellbeing, resource conservation and reduced lifecycle costs. For avoiding mobility demand and shifting to active modes and public transport, government and spatial (city) planners are the key actors and help in avoiding service demand and emissions. For the industry sector to create a symbiotic relationship among the firms, facilitating the role of the IS and EIP operators and policy support from the government is crucial. Strengthening of the financial institution with innovative financial support mechanism is also needed. The role of research organizations is also important to develop a knowledge base for the possible use of by-products and technological development.
Behavioral change is often constrained by non-behavioral factors like policy, education, and infrastructure. Especially high upfront costs of low carbon options, inadequate infrastructure, incorrect understanding or low awareness regarding sustainable options and socio-cultural factors define behavioral choices. A better understanding of the costs and benefits of these solutions could facilitate higher adoption.
Implementation of mitigation actions that can help stabilize global warming at 1.5 • C has multiple synergies across a range of sustainable development dimensions. Not surprisingly, we find a number of actions across all three sectors that offer opportunities to deliver simultaneous benefits for climate change and SDGs. Clean cooking fuels, investments in public transport and sustainable and green buildings are examples of such solutions. All avoid actions and most of the shift actions entail reduced consumption while most of the improve actions result in responsible consumption therefore aligning with SDG 12.
Most of these demand side mitigation actions showed positive linkages with SDGs 7, 9 and 11. The recent upscaling of digitalization offers sizeable opportunities for industrial innovation. Initiatives such as active transport, passive design, cleaner cooking fuels could offer the same level of service with additional synergies (co-benefits) across several SDGs.
The review brings out several studies that point to the need for looking at the longer term in a sustainability context. A case in point is green buildings where performance could possibly decline over time and this can only be reviewed through long-term monitoring studies (Zaid et al 2017, Ge et al 2018. Similarly, the rebound effect of energy efficiency occurs over time and should be an important consideration in designing energy policies. Similarly, switching to cleaner cooking fuels is a special case where the transformation needs to be viewed beyond its energy benefits as it delivers significantly wider benefits for gender equity and development in low-and middle-income countries.
Although the review was limited to only one database, Scopus, findings present evidence of the diversity of actions including emerging innovations that reduce energy consumption, generate cost savings and offer sizeable opportunities to achieve targeted benefits. Digitalization through ICT and IoT has emerged as a significant enabler of various demand side mitigation actions. There are case studies to show that technologies such as ridesharing in transportation, enhanced efficiency to smart buildings, appliances and smart meters have enhanced service delivery, while reducing material consumption and CO 2 emissions in the app sector are some examples.
Several actions including public transit, industry clusters or district heating are enabled by integrating into existing planning regulations. Spatial planners (e.g. urban planners and architects) through planning interventions including zoning, land use changes, passive design strategies have a significant influence to shape urban form to a more sustainable and climate responsive one. Emerging developments in ICT provide for a significant opportunity for industrial innovation in devices and services.
One important finding of this study is that demand side actions cannot be effective enough in isolation. For optimizing the mitigation actions, the interconnectedness across sectors is important. Demand side mitigation actions in the transport sector are highly dependent on city planning and infrastructure and the construction sector is an integral part of the building sector. However, for the industry, it is found that some demand side actions of other sectors have some tradeoff in industrial production. In order to achieve an overall emission abatement, it is important to develop a cross-sectoral framework. Similarly, technological solutions can be more effective if complemented with sustainable behavioral responses of end users.
The success of many of the demand side actions depends on the willingness of people to modify their behavior, which is rooted in complex economic, social and cultural circumstances besides the design of built structures. Studies have also highlighted the challenges and complexities involved in switching to a sustainable lifestyle (Konrad 2015). In many cases, it is a lack of awareness or incorrect understanding and these can be corrected by enhancing awareness, through better reporting and feedback on consumption. Organizations could play a significant role in shaping behavior by monitoring and understanding consumption habits to set a sustainability culture, through regulations and targets such as switching off appliances when not in use, reducing personal electricity consumption, bike to work, etc. Educational institutions such as schools and universities have a significant role to play by incorporating sustainability into curricula, empowering students to make sustainable lifestyle choices, and through research and living labs, offer a testbed for potential solutions that can co-deliver mitigation and wellbeing within the premises but also at a wider spatial scale to neighborhood and the city and temporally by shaping mindsets of occupants toward responsible consumption.
We have adopted a semi-systematic but comprehensive review method where we have systematically searched and screened the existing huge body of literature using the Scopus database and only literature in the English language. This is because performing a full systematic review is time-consuming, resourceintensive, focuses on a narrow range of questions and also prevents in-depth analysis . However, studies also suggest that the use of a few more databases ensures adequate coverage (Bramer et al 2017). We are aware of this limitation in our study but we feel that this study has attempted to deliver a comprehensive in-depth analysis of demand side mitigation actions and SDGs, and actors' role. Most studies identify benefits that could indirectly contribute to SDGs, while few studies directly establish linkages with SDGs, especially at the target level covering the study period. The findings from several case studies are context specific. The study recognizes that SDG relationships are complex and often not easily quantifiable, non-linear and vary across time.
Few studies look at impacts of mitigation pathways, policies and actions on reducing poverty, affordability of technologies and equity impacts. Future studies could take a deeper look into understanding these complex interlinkages, especially spatial and temporal trade-offs between mitigation and SDGs, and maybe also in many other country contexts.

Data availability statement
All data that support the findings of this study are included within the article (and any supplementary files). Avoiding demand for industrial products and services is driven by the final consumer, and is therefore considered as an additional search term.

Annex B
• Title: Presence of specific sector names or their synonyms. Mention of mitigation actions that the authors later identified as ASI category • Abstract: In addition to the criteria for title selection, mention of implementation of the actions, hint of cross-sectoral linkages or deep decarbonization or zero emission • Full text: At this level the selection criteria is narrowed down to articles that aligned with the research question. These would include ASI mitigation actions for transport, industry and buildings, evidence of implementation and including mention of sustainable development or sustainable development goals or sustainability. Experimental, modeling and lab-based studies were excluded from the final analysis.
Given the diversity in the demand among these three sectors, apart from these general criteria for selection, we have also followed some sector-specific criteria.
Transport: While screening through the titles, specific mention of supply chain, manufacturing or related words, vehicle design, microgrids integration of electric vehicles, distribution grids in presence of EVs charging stations are excluded. During screening through abstracts along with all those mentioned in title screening and power flow analysis, studies regarding life-cycle of batteries assessment, implementing dual fuel, EV batteries recycling, rail-line or pavement or road construction materials or maintenance, choice between sustainable mobility plans by policy makers are excluded. At the third stage of screening we excluded studies which did not include sustainability/sustainable development/ SDGs.
Industry: Title screening included articles which mention any of the eight major CO 2 -emitting industries mentioned in IPCC AR5, namely, iron and steel, cement, chemical, pulp and paper, non-ferrous metal, food processing, textile and leather and mining or manufacturing industry as a whole. Articles on all other industries are excluded. At the abstract screening stage, studies relating to business models for mitigation actions, life cycle analysis (LCA) are excluded. While screening full text studies, we excluded studies that did not specify any specific demand side actions. Studies regarding SME other than manufacturing industry are excluded. Industrial park design and business sustainability are also excluded. We have excluded experimental studies particularly those which do not have any implementation in real world settings.
Additional literature has been selected from the results of a separate search in Google Scholar for the industrial linkage of some specific demand-side actions such as the change in demand for pulp and paper industry with increase in electronic newspapers, reduced textile demands from increase in clothing longevity, reuse and recycling of textiles, and altered cement demands from the use of wood ash. These additional literatures are included based on the authors' knowledge from grey literature, news items of emerging actions on demand reduction at the end-user level and as we were looking for very specific actions, in Google Scholar we were getting more explicit targeted results.
Building: Title screening involved screening of papers which included studies specific to buildings sector and mitigation actions on the demand side. This would automatically exclude studies on other sectors such as transport and industry, as well as supply side actions. At the abstract screening stage, a number of studies were found to focus on construction materials. These were screened out since these would involve substitution with other materials and therefore are addressed in the industry sector. Studies involving economic dimensions including costs and benefits of mitigation actions and those relating to the wider context such as modeled future urban scenarios were not considered. Similarly, papers that included theoretical studies except for literature reviews, such as desk research, lab experiments, simulations or modeling exercises and did not offer evidence of implementation were excluded. For example, new and innovative experiments involving alternate materials, thermal properties, new designs and approaches for construction and simulations of energy consumption were not considered. An additional search was conducted to include articles specifically on clean cooking, since this is an important dimension and was not captured in the original Scopus search.