Accounting for gender in climate policy advice: adapting a gender impact assessment tool to issues of climate change

ABSTRACT

This article discusses how gender impact assessment (GIA) can be adapted and utilised to better include gender aspects in climate policy advice. The article explains the re-modelling of the equality governance tool gender impact assessment in the context of climate change and adaptation policies, programmes and projects on behalf of the German Environment Agency. While the focus of the tool remains on women and men, it operates with an intersectional and non-binary gender+ concept and understanding of sex and gender relations. The main novelties of the tool are its evidence-based educational elements and that it encompasses seven gender dimensions addressing the main areas of life that play critical roles in producing and reproducing gender disparities. This allows for the identification of how climate policies interact with gender relations and can make substantial contributions to gender equality, as well as of areas in which addressing gender issues could render climate policies more effective. Drawing on the case of the GIA tool in Germany, we argue that core insights gained during the tool development and testing can be applied to gender mainstreaming strategies of climate policies and interventions in other jurisdictions, as well as holding potential for adoption in other policy domains.

KEYWORDS:

• Sex and gender analysis
• policy analysis
• climate protection
• climate adaptation
• equality governance
• gender dimensions

1. Introduction

The gendered impacts of climate change have long been highlighted in feminist research and gender studies (Dankelman 2010; Çağlar et al. 2012; Alston and Whittenbury 2013). It is also increasingly acknowledged in international agreements, such as the 2030 Agenda for Sustainable Development, where to ‘achieve gender equality and empower all women and girls’ is identified as the cross-cutting goal number 5 (UN SDG 2020). The documents of the three Rio Conventions (United Nations Framework Convention on Climate Change (UNFCCC), Biodiversity Convention and UN Convention to Combat Desertification) contain an ever-growing number of references to gender. These high-level agreements, including their gender mandates, are supposed to be implemented through the national laws and policies of the signatory states. The Paris Agreement (a key agreement under the UNFCCC dealing with greenhouse gas emissions mitigation, adaptation and finance, signed in 2016), its implementation guidelines (the so-called ‘Katowice climate package’ from 2018) and Gender Action Plans (for current plan see UNFCCC 2019) in particular are an attempt to shape gender-responsive climate policies in a structured fashion. Since climate change is generally perceived as being ‘gender neutral’ (Bonewit and Shreeves 2015), policies internationally are usually rolled out in a non-gender-responsive fashion (EIGE 2012). As these policies not only affect issues of protection, mitigation and adaptation, but also gender relations, recommendations such as those of the UNFCCC call upon governments to promote and implement gender impact assessments (GIA) of climate policies, programmes and projects. For instance, states are called upon to: ‘Enhance capacity-building for governments and other relevant stakeholders to collect, analyse and apply sex-disaggregated data and gender analysis in the context of climate change, where applicable’ (UNFCCC 2019, p. 5).

In 2016, in an attempt to increase the use of GIA and the quality of analysis results, the German Environment Agency (‘Umweltbundesamt’ (UBA)) commissioned the research project that provides the basis for this article. The project, which ran until 2019, pertained to the UBA’s gender mainstreaming as well as the UNFCCC’s gender mandate. Focused on the Global North and Germany specifically due to the mandate of the UBA, it investigated how a gender equality perspective could be introduced and strengthened in UBA’s research and policy advice on climate policy and thus improve overall analysis of climate change and adaptation measures and policies (Spitzner et al. 2020). The research project was carried out by a consortium of the Wuppertal Institute, GenderCC and ISOE. The authors of this article assumed distinct roles in the research process, with one of us having commissioned the research project on behalf of the UBA (Sauer) and the other having participated as a member of the research team of the Institute for Social-Ecologial Research (ISOE) (Stieß). A main objective of the research project was to create a comprehensive knowledge base on climate and gender-related publications (peer review articles in journals, research reports and books published since 2010 that referred to the situation in the countries of the Global North). A second objective was to design a specific GIA for climate policy and programme-making that can also be applied to the evaluation of research projects commissioned by the UBA (Spitzner et al. 2020).¹ Drawing on the experience of this project, in this article we present the core insights gained during the tool development and testing, with a view to contextualising the findings of the research project in the broader discourse of impact assessment. We argue that GIA should apply an intersectional gender+ approach (these core concepts are described in further detail in section three of the article), accounting for diversity within sex and gender disparities, while not losing focus on gender equality governance. The gender dimensions that were established as part of the GIA tool development, and validated against this body of research, provide a novel and beneficial conceptual and analytical framework for GIA and foster conceptual and instrumental learning. The development of the GIA tool is situated in the institutional context of UBA. The focus of application is on policy and research project analysis in environmental departmental research rather than on explicitly regulatory use. Nevertheless, we argue that the insights of the research project may also prove useful beyond climate policies, programmes and projects, if transferred to other policy areas and non-governmental contexts. Furthermore, while acknowledging the specific country and Global North context of this particular research project and GIA tool development, the lessons learnt may hold relevance for other jurisdictional and global contexts where actors are working towards greater integration of gender considerations in climate policies, programmes and projects.

Our article proceeds as follows. In the next section, we provide some background knowledge on the evolution of GIA in order to contextualise GIA tools within gender mainstreaming (including associated quality criteria for tools) and impact assessment (IA) development. In the third and main section of the article, we introduce the climate GIA developed as part of the project, and its gender dimensions as the analytical core of the tool in more detail. We then present insights from a workshop in which these dimensions were tested with users. The final shape of the tool, including the gender+ concept, is laid out at the end of section three. In section four, we discuss how the new climate GIA adheres to gender mainstreaming principles (including feminist intersectionality and gender+ concepts), and how the tool fits the IA implementation context while also allowing for conceptual learning about gender. We share lessons learnt from the climate GIA refinement and testing process and consider advantages and disadvantages of an educational, evidence-based and therefore rather comprehensive tool. We also briefly reflect upon the (possible) transferability of the gender dimensions to other policy areas and the contextual factors necessary for successful implementation. In the fifth and final section, we wrap up the discussion and outline our main findings pertaining to how and under what conditions the climate GIA can be suitably implemented.

2. Gender impact assessment as a public equality governance tool

IA is ‘the process of identifying the future consequences of a current or proposed action’ (IAIA 2020) and predominantly tool-based. Policy GIA tools have been created since the inception of gender mainstreaming at the Fourth United Nations’ World Conference on Women in Beijing in 1995, in order to help implement this international, overarching strategy for realising gender equality between women and men and combating discrimination. Gender mainstreaming requires governments to assess ‘the implications for women and men of any planned action, including legislation, policies or programmes, in all areas and at all levels’ (UNGA 1997, p. 24).

GIA tools operationalise the integration of a gender perspective via ex-ante policy, programme or project IA in public administration. Implementation suitability therefore requires GIA tools to match quality requirements with regard to IA and gender mainstreaming (Sauer 2018). In order for tools to meet the quality requirements of gender mainstreaming, they (among other factors) must be based firmly on feminist concepts. Precisely which feminist concepts are rendered relevant is, however, a matter of dispute. The theoretical basis for the first generation of GIA tools was a liberal feminism of difference, while recent debates foster poststructuralist or postmodern feminist and even queer feminist approaches. The relevance of intersectionality, largely undisputed since Crenshaw’s (1989) interventions, seems indispensable for the gender assessments of today (see also Levac et al. 2021, this volume). The primary goal of achieving gender equality cannot be separated from intersectional issues and social justice at large (Hankivsky et al. 2019).

Partially due to a lack of accountability and oversight for GIA application, but also due to a lack of understanding of feminist concepts and failure to relate them to the subjects of IA investigation at hand, the general practice of GIA has proven to be difficult in terms of practitioners’ receptivity and the institutionalisation of the instrument (see also Götzmann and Bainton 2021, this volume). Fully executed and documented policy GIAs are hard to find in public administration (Lewalter 2013; Sauer 2018; Veit 2010; see also Maduekwe and Factor 2021, this volume). Feminist criticism was quick to point to lagging GIA use as confirmation that GIA takes an integrationist approach and as such is void of a transformative potential, due to a lack of feminist concepts within the tools, a lack of their understanding (i.e. by bureaucrats) and a lack of conducive environment (in structures and institutions) (see Sauer 2018, p. 37–45).

Non-feminist critiques about the (non-)usability of GIA tools are also present. Current and potential instrument users in public administration call instrument design into question for its theoretical complexity and for being insufficiently concrete, which they argue renders GIA instruments unfit for implementation in specific policy areas. Current practices of ex-ante policy or programme advising through IA in public administration remain reductive and fail to include a multiplicity of important perspectives and democratic obligations, i.e. a gender equality perspective (see Sauer 2018, p. 33–63). Even on a project level, the need for IA practitioners to integrate gender equality concerns into social impact assessment (SIA) might be ‘gaining traction in the literature’ (Adusei-Asante and Pelden 2018, p. 1), but it is not materialising to the same extent in practice (see also Götzmann and Bainton 2021, and Hill et al. 2021, this volume).² Our hypothesis is that this is also due to a lack of educational elements and learning possibilities in these tools.³ These learning possibilities are described by Hertin et al. (2009) as:

Conceptual learning: when knowledge ‘enlightens’ policy makers by slowly feeding new information, ideas, and perspectives into the policy system, challenging existing beliefs and opening up new opportunities for policy change. … Instrumental learning: when knowledge directly informs concrete decisions by providing specific information on the design of policies. (p. 1187)

The absence of conceptual learning also becomes evident in IA discourse: Even 25 years after the inception of gender mainstreaming, current policy or environmental IA handbooks or methodological overview articles usually devote no more than a small paragraph or a passing mention (if at all) to gender issues. If they do, they usually frame women as a vulnerable group, only ask for sex-disaggregated data in i.e. community/population assessments (Dunlop and Radaelli 2016; Geneletti 2016; Therivel and Wood 2017; Carroll et al. 2020), or are ‘othering’ the tool by being sceptical regarding IA fit, purpose and outcome of GIA (Gains 2016). Due to their lack of a gender perspective, mainstream IA tools reproduce rather than transform the gender hierarchies and imbalances in institutional practice. An evidence-based conceptual re-design was arguably required to put GIA (back) on the IA map.

3. Gender impact assessment tool for climate policies

The research team took the German regulatory GIA baseline tool (BMFSFJ 2007) as the starting point for our project. The predecessor of this instrument was developed originally by the ISOE, in a project for and with the UBA and the German Federal Ministry for Environment, Nature Conservation and Nuclear Safety (BMU) in 2002, and emulated environmental impact assessment (EIA). This early BMU GIA was successfully tested in the case of radiation (Hayn and Schultz 2002); however, it was not routinely applied afterwards. In a process similar to the development process of SIA, which initially was also modelled closely on EIA, the UBA contracted the development of a new climate GIA. To assess gender issues properly, the UBA needed a GIA that matched specific policy areas and application contexts. The UBA wanted a usable, context-specific, validated and tested instrument, and the research team responded by rendering previous instruments (BMFSFJ 2007) applicable to climate research project, policy and programme design for the working contexts of the UBA and BMU.

As a first step to tool development, an inventory of GIAs in various fields of climate policy action was carried out. The international search was conducted in English and found examples from various European countries, Canada and the Republic of Korea (for a policy example from West Africa in the context of the energy sector see Maduekwe and Factor 2021, this volume). Methods of impact analysis from other areas that address gender-relevant issues, e.g. technology assessment or public health research, were also evaluated. In order to adapt the GIA tool to the issues of climate policy, a theoretical basis founded on the intersectional and non-binary gender+ concept as well as a new conceptual framework based on so-called gender dimensions were established and validated against the empirical findings from gender-reflective research on climate change (Röhr et al. 2018). In the end, the newly designed tool was tested in workshops with potential users, adapted in accordance with the results of these workshops, and subjected to gender experts’ peer review, before it was presented in a concluding feedback workshop to UBA and BMU experts and academic scholars. All of these steps fed into the final tool content (Spitzner et al. 2020).

3.1 Tool methodology: gender dimensions

With the GIA tool, climate policy projects such as strategies, programmes, projects or concrete measures should be evaluated and optimised with regard to gender equality, so that they are better aligned with the needs of their target groups (Spitzner et al. 2020). The gender analysis ensures that the action that is taken is tailored to meet the needs of women, men and, where relevant, diverse genders. At the same time, it can open up a broader spectrum of solutions and improve the impact of climate change and adaptation policy and programme-making, in projects and in strategies.

The basic idea of the project was to build the GIA tool upon analytical categories that account for societal power relations (causes, structural inequalities and manifestations) between the genders, the so-called gender dimensions, as developed in Spitzner et al. (2020), thereby drawing on different theoretical and conceptual considerations.⁴ These gender dimensions point to central areas or fields of societal structuring through which hierarchical gender relationships are established, maintained and reproduced (Verloo and Roggeband 1996; Verloo and Lombardo 2007). These interconnected structures include the gender-based distribution of opportunities, burdens and risks in a society as well as hierarchical norms, biased perceptions and mechanisms of access to and exclusion from institutions and organisations. Drawing on the seminal work of Verloo and Roggeband and on insights from GIAs in the field of environmental, research and transport policy (cf. Schultz et al. 2001; Turner et al. 2006; Alber and Röhr 2011), the conceptual framework of the tool was enhanced.

Finally, six gender dimensions and a seventh cross-sectional dimension (‘symbolic order’) were devised, representing areas of life in which a sex and or gender-specificity can still be discerned. These gender dimensions describe the social areas of the care economy and caring work, employment, shaping of public infrastructures, institutionalised rationalities, participation in decision-making and physical integrity (see Table 1).

Table 1. Overview of gender dimensions

Gender dimensions
Symbolic order: hierarchisations, attributions and positions of meaning, gender-hierarchical narratives and modernisation strategies (cross-section dimension)	Care economy/care work: attribution, significance, distribution, instrumentalisation in societies’ economies
	Market driven/labour economy: horizontal and vertical segregation, economic job valuation, poverty, property, financial capabilities
	Public resources/infrastructures: provision, alignment, prioritisation, accessibility, usability
	Institutionalised androcentrism/power of definition: masculinity models set as the standard and benchmark in subject-/action-field-specific rationalities, problem perceptions, methods etc., institutionalised content-related-cognitive gender hierarchisations
	Shaping power at actor’s level: Participation and consideration of gender expertise in decision-making in science, technology and politics
	Body/health/safety/privacy (intimacy): social organisation of sexuality, health, freedom from violence, privacy, sexual self-determination

(Spitzner et al. 2020, p. 31, English translation adapted by authors).

In a second step, the generic model of gender dimensions was validated and adapted to topics of climate change and climate policy, using a comprehensive literature review of climate- and gender-related publications that highlights the differences, inequalities and power imbalances between women and men (and other genders) with respect to climate policies and programmes (Röhr et al. 2018). The literature review was conducted in the spring of 2017 and was essentially limited to peer review articles in journals, research reports and books published since 2010 that referred to the situation in the countries of the Global North. Findings from this gender-reflective research were available in almost all fields of climate policy, albeit to a variable extent. These findings provide material that illustrates gender aspects of climate policies in various societal, politically accountable areas of life (gender dimensions): E.g. as women tend to have lower incomes than men, they are more susceptible to rising energy prices. Basic findings were included in the description of the gender dimensions, providing empirical evidence-based knowledge about gendered inequalities. This knowledge helps users identify the potential effects of policies and measures on gender relationships and generate opportunities for political action.

3.2 Tool testing

The prototype of the GIA guidance was tested with UBA experts in December 2017. The main objective was to find out whether the guidance was intelligible to its intended users and could provide them meaningful insights into gender aspects of policy-making. Two ongoing UBA projects were selected as cases for the tool testing, one from the field of greenhouse gas mitigation and one from adaptation to climate change impacts. Both examples have a direct gender relevance. The first case addressed scenarios for reducing the degree of motorisation as a contribution to a compact mixed-function city; the second made recommendations for behavioural heat protection in the context of heat action plans. For both test cases, the project team collected and summarised evidence on gender aspects from the literature review in short inputs. The UBA experts then analysed and structured the two test cases with the help of the gender dimensions and the GIA test questions.

3.2.1 Test Case 1: Scenarios for reducing the degree of motorisation as a contribution to a compact mixed-function city

The test example ‘The compact, function-mixed city’ (Penn-Brussel 2017) envisages, among other things, a significant reduction in the motorisation grade in cities. In the workshop, potential impacts of this scenario on the areas ‘care economy/care work’, ‘employment economy’ and ‘public resources/infrastructure’ were examined in greater depth. The analysis shows that changes in the motorisation grade have a high gender relevance. An example is the attribution and organisation of care work (gender dimension ‘care economy/care work’). Escorting relatives in need of care or relatives with restricted mobility, for example during visits to the doctor, must be organised without having or owning a car. If car sharing is not available, these trips must be organised with public transport, which may result in additional time expenditure. On the other hand, traffic calming and safe paths for walking and cycling can help children to make their way to school on their own and reduce the amount of care work involved. Furthermore, intersections of gender and income need to be considered, because households with low income often live in settlements with poor access to public transport.

A reduced motorisation grade also has an effect on the dimension of ‘labour economics’. The lower demand for cars leads to changes in the gender-specific division of labour, since the reduced demand entails a decline in jobs in the male-dominated sectors of vehicle production. On the other hand, more jobs will be needed for maintenance and cleaning, for example of public transport or repair of bicycles. This creates new qualification profiles and job opportunities. Overall, therefore, a lower motorisation grade offers the opportunity to reduce the existing segregation of employment. A revival of the local economy also generates new opportunities for securing one’s own livelihood and could possibly create additional jobs for salespersons or suppliers.

3.2.2 Test Case 2: Recommendations for behavioural heat protection in the context of heat action plans

The global warming caused by climate change and the associated increase in the frequency and duration of heat events can have a severe impact on people. Heat stress, exhaustion and heat strokes can have fatal consequences. Statistical data show that mortality rates among women during heat waves at different ages are sometimes twice as high as those of men (cf. D’Ippoliti et al. 2010; Filial et al. 2011). From a gender perspective, it has been argued that the differences in mortality are resulting from an interplay of physiological effects and social factors, like the gendered responsibility for care and nursing in families. The higher mortality of women is mainly explained as a consequence of the higher proportion of women who live alone and do not rely on a partner’s care (D’Ippoliti et al. 2010).

The object of the testing was the UBA information brochure ‘Der Hitzeknigge’ (The Heat Etiquette). The brochure is designed as an information tool for adapting behaviour during heat events (Grothmann and Becker n.d.) in the context of heat action plans. In the workshop, the brochure was analysed for relevant gender aspects in several areas of life. The focus of the discussion was on the areas of life ‘care economy/care work’, ‘symbolic order’ and ‘body/health/security’. The testing showed a high degree of gender relevance. In the life area ‘body/health/safety’, the different risk awareness of men and women is an important factor that must be considered when communicating recommendations for behaviour-oriented heat protection. Stressing the principle of personal precaution, the brochure addresses both men and women, so that a gender-specific attribution of unpaid care work is avoided. The adapted climate GIA encouraged the participants to consider the different life situations of women and men, which are also shaped by age and the household type: In particular, elderly people living alone are most vulnerable to the impact of heat waves. The workshop participants found this guidance very helpful because it provided new insights for improving the dissemination of the brochure by choosing suitable information channels and multipliers.

The testing workshop demonstrated that the GIA guidance and the gender dimensions helped users to identify gender aspects of climate policies. It was evident that the GIA test users gained new insights and found the tool structure as well as the catalogue of questions appropriate. The testing also provided important suggestions about how to improve the practical suitability of the instrument and facilitate its application. These suggested changes mainly pertained to:

• the need for a concise explanation of the pertinent sex and gender concepts as well as a more in-depth explanation of the gender dimensions in a preamble, in which each gender dimension would be illustrated by a short introduction of relevant research and the underlying evidence-base (data, main results); and

• the need for an explanation of the potential added benefit for climate protection and adaption, to be illustrated with a few research examples with gender and climate relevance, in order to enhance the tool’s suitability for implementation and users’ willingness to use the tool.

These suggestions were taken up. The questionnaire was rearranged and shortened to avoid an overlap of questions; all questions were re-formulated as closed questions and entail the obligation for providing evidence-based arguments for the answers given. The structural presentation and layout of the tool description were improved in order to facilitate its use.

The updated tool version was then sent to gender experts for comment and reshaped again according to their suggestions. The revised GIA instrument was finally presented to a wider interested public and to public servants of the UBA and BMU in a concluding public workshop of the UBA in 2018. In this workshop, practical exercises with examples (from the areas of heat protection and cycling promotion) were presented again to the participants, in order to explore how the catalogue of questions could be used to make clear the significance of gender for climate policy.

3.3 Final tool shape and content

The re-designed tool incorporates a more fluid understanding of gender by expanding on the gender+ concept as it was originally developed in the EU QUING project (Walby et al. 2009; Verloo and Walby 2010). Gender+ originally focused on ‘a plus’ in intersectionality, but was still based on a binary understanding of gender. Gender+ in the climate GIA addresses sex/gender not as homogeneous or as strictly binary, but as categories which are to be analysed context-specifically and in interplay with other factors of social inequality or social marginalisation (see also Levac et al. 2021, this volume). The main assessment focus remains on women and men, but users of the tool are encouraged to also take transgender, intersex and non-binary people into account as well as to consider non-heterosexual orientations in the analysis (according to context, assessment focus and available data). Regularly, gender+ should be extended to such groups, when gender dimensions referring to known areas of discrimination like access to labour market economy, public resources and infrastructures or body, health, safety and privacy (intimacy) play a role in the assessment.⁵

Starting out from binary and (when relevant) such non-binary sex/gender+ considerations, interactions with other characteristics of inequality such as age, income, education, physical constitution, cultural background and so forth, are to be taken into account.⁶

The availability of data is a crucial question for (binary) GIA in general and intersectional and gender+ based GIA in particular, which cannot be explored in this article. Spitzner et al. (2020) already highlighted data gaps and suggested, how climate policy-making in Germany can address them better. For example, statistical surveys should systematically include intersectional indicators (like income, household type, caring responsibilities or age). Qualitative data such as community-based case studies provide valuable sources of information where statistical data is not available. Finally, data can and should ideally also be generated in the course of the assessment, where data gaps became evident.

Methodologically, the effectiveness of a specific climate intervention in terms of gender equality is examined in a procedural, iterative approach. In a first step, the effects of the policy/programme/project to be investigated on the hierarchical relationships between the sexes/genders are examined. In a second step, further socio-economically differentiated features are introduced into the analysis in order, for example, to examine the impact of a measure with regard to discrimination and exclusion along characteristics such as age, income, religion and so forth. It should be noted that no previously defined hierarchies of inequalities can be assumed (see also Reynolds 2021, this volume). Various factors that generate inequality must be considered in a case- and context-specific manner; the categories and their significance must be worked out in the research process itself. No claim is made to include all categories that are involved in the constitution of inequalities (social, economic, health, etc.); rather, the focus should be on those that have the highest relevance for the respective field of action. Decisions about how to select and demarcate the relevant differentiation features are to be made on the basis of the current state of the scientific debate and evidence on gender in the various fields of action of climate policy.

The thematic structure of the GIA is still based on areas of life (gender dimensions), which are described and illustrated with examples. They serve as a kind of search matrix for the identification and analysis of equality-relevant effects in climate-relevant fields of action. Concrete test questions are formulated for each area of life.

The gender equality-oriented impact assessment is carried out in two steps (See Table 2). With the help of a relevance test, it is first examined whether a climate intervention may have gender equality impacts or not. If gender equality relevance is established and elucidated in a brief justifying statement, the main assessment is carried out. If gender relevance is not established, a statement with the according (negative) evidence must be laid out too.

Table 2. Overview of the two-step GIA procedure

Pre-test of relevance
Initial situation	Subject, measures and evidence-base of intervention? Which intervention is intended (ex-ante) or is to be investigated in parallel or ex-post?
Determination of relevance for gender equality	Is gender+ directly or indirectly affected by the intervention or individual parts thereof? Which areas of life are affected by the intervention? Are there indications that the intervention or parts of it have different impacts on gender+?
Assessment	Determination of gender relevance of the intervention (yes/no) with short written justification
Main examination
Initial situation	To which facts does the intervention relate? What are the technical objectives of the intervention and how are they justified? What data (statistics, studies, etc.) and findings form the basis of the intervention?
Analysis of the project	Which measures are planned to achieve the objectives or have been defined to achieve the objectives (incl. indicators)? Which actors are (or have been) involved in the design of the intervention? Which alternatives, variants and accompanying measures have been examined and with what result?
Analysis of impacts on gender equality	Which groups of people are directly and indirectly involved or affected? What equality-relevant effects does the project have on different life situations, interests and everyday practices in relation to gender+? Are there any prominent discrimination or disadvantages due to other characteristics of social inequality of the affected groups of people (e.g., age, social situation, origin, etc.) that have an influence on the impact of the intervention? Which relevant actors should be involved, in what form and at what point in time, in order to take gender expertise and a gender perspective into account? How must the measures for achieving the objectives of the intervention be adapted (including indicators and actors) or which technical alternatives or accompanying measures must be introduced for gender responsiveness?
Evaluation	Written report

(Spitzner et al. 2020, p. 35, English translation adapted by authors).

The main assessment consists of an in-depth examination. It comprises both qualitative and quantitative procedures and is based on the questions posed by the relevance test. In the main appraisal, a more detailed analysis of the impacts on the areas of life (gender dimensions) is carried out. Part of this is a more detailed description of the genders directly and indirectly affected by the intervention and the consideration of other characteristics of social differentiation, sexual orientation and gender identity (gender+). As a result, the assessment determines the impacts that the climate project/programme/policy will have on fields of action in gender equality policy and suggests concrete opportunities and recommendations for improving gender equality.

4. Discussion

The tool development took place in Germany, but it was based on a broad body of (mainly English-language) research pertaining to the Global North. As such, the concluding discussion pertains primarily to GIA design and implementation internationally in Western countries and comparable implementation contexts such as the German environmental bureaucracy. Without presuming transferability to contexts beyond this, the lessons learnt may, however, also hold relevance for other jurisdictional and global contexts where actors are working towards greater integration of sex and gender considerations in climate governance. Given the pressing need for action on gender and climate globally, we urge for further research and practice development in this regard, with a view to developing locally-relevant solutions for strengthening attention to gender in climate policies, programmes and projects. Our discussion will also reflect on possible transferability to other policy domains and what would be needed to investigate such potential.

4.1 Impact assessment and gender mainstreaming fitness

The climate GIA is an explicit IA tool (Sauer 2018). Its implementation is mandated by the gender mainstreaming mandate, the cross-cutting nature of SDG 5, especially at the junction of SGD 13 ‘climate action’, and finally the Paris Agreement, its implementation guidelines and UNFCCC Gender Action Plan (UNFCCC 2019). The new climate GIA now allows for conceptual learning about feminist gender+ approaches as well as instrumental learning about sex/gender and climate issues (Hertin et al. 2009). For instance, an exemplary and comprehensive GIA on measures to improve the cycling infrastructure in municipalities for less carbon intense transport would provide more insights into gender issues around cycling. As a previous study in Australia demonstrated: On average women tend to cycle shorter distances with lower speed taking stops due to shopping and care responsibilities, whereas men often prefer to cycle longer distances and faster mostly from/to work; both genders prefer bicycle lanes, but women were more concerned with their safety (Heesch et al. 2012). These insights would need to be validated for Germany and if applicable could then be transferred to local contexts to inform bicycle lane infrastructure planning.

IA practitioners should be interested in such in-depth evidence, since they consider themselves largely to be ‘fact oriented and “rational” actors (that is, engaging mostly in learning and justificatory knowledge use), while accusing other players of using research in a biased, strategic, or symbolic way’ (Hertin et al. 2009, p. 1198). A tool with a strong evidence base, thus elevates trust and decreases suspicion of political usage within the IA community and analysts. By definition, political IA usage occurs when an ‘appraisal is not based on the ideals of evidence-based policymaking, but is used by affected social, economic, and political interests’ (Dunlop et al. 2012, p. 27). Gender issues have been and often still are seen, as political interest management and as being strategically biased towards women (for the realm of regulatory IA cf. Gains 2016).⁷ Seen in that light, the climate GIA is an unabashedly political gender mainstreaming tool (Sauer 2018) that fosters gender equality as a constitutional and therefore political objective. Looking down on GIA as political ‘interest management …, while ignoring the per se political character of all policy IA and general messiness’ (Sauer 2018, p. 67) will now be harder due to the epistemic strength of the GIA tool combined with its powerful critical, reflective and educative elements.

The feminist theories, gender dimensions and sex-disaggregated data in which the climate GIA is deeply rooted intentionally bring ‘explicit political conflict within the economic analysis’ (Dunlop et al. 2012, p. 27) to the discussion table. They also offer more complex data and different interpretations rather than relying on exclusively androcentric data and bias-unaware rational choice theory. The existence of ‘competing rationalities’ and their ‘blurred boundaries between science and policy, facts and norms’ (Hertin et al. 2009, p. 1198) are now widely acknowledged in the IA community (see also Götzmann and Bainton 2021, this volume). By advancing gender knowledge ‘to attain political objectives’ (Hertin et al. 2009, p. 1187) of the SDGs and the climate accords, climate GIA is now also urging policy IA to acknowledge its role in the realisation of gender equality as a norm. The climate GIA is an analytical tool, providing a more comprehensive inter- and transdisciplinary knowledge base against which climate policy measures can be evaluated.⁸ This stimulates the development of a broader spectrum of solutions that allows for a more precise consideration of the needs and living conditions of the addressees. Climate policy strategies, programmes, projects and measures can thus be designed more accurately and blind spots or unintended consequences can be identified and avoided at an early stage. In doing so, it adds accountability to the reporting requirements under the climate accords (Paris Agreement) with regard to gender responsiveness and achieves better climate outcomes. The downside of such a research-based, comprehensive, and consequently comparatively long tool might be that it requires relatively high maintenance, in that its research base should be updated at regular intervals and the gender dimensions (possibly) adjusted. This maintenance could be counterproductive for tool uptake, but could also, in the long run, be advantageous: The re-occurring issue of quality management in tool content and design will draw regular attention to climate GIA in the toolbox and prompt a regular IA tool evaluation process that leaves room for additional improvements and updates.

4.2 Tool design and transferability of the gender dimensions

As its main novelty, the climate GIA tool is based on so called ‘gender dimensions’, derived from the main areas of differences and power inequality between women and men as found in climate change and adaptation research for the Global North. The gender dimensions help structure the analysis of gender aspects and provide knowledge and create awareness among users who are not (yet) familiar with gender issues in their field of work. It is interesting to note that many indicators of gender indexes used in the Global North show a large overlap with the gender dimensions used in climate GIA, indicating that the inequalities and disparities women are experiencing in Western societies with regard to climate-relevant issues are also comparable to other policy fields. For instance, the Gender Equality Index (GEI) of the European Union (EIGE 2019) uses the following seven compound indicators (the climate GIA gender dimension with the most overlap are shown in parentheses): work (market driven/labour economy), money (market driven/labour economy), knowledge (public resources/infrastructures), time (care economy/care work), power (shaping power at actor’s level), health (body/health/safety/privacy – intimacy) and violence (body/health/safety/privacy – intimacy). Absent from most index indicators, however, are overarching questions of how the hierarchisation of women and men, female and male – conceptualised as ‘androcentrism’ – manifests in each category and how it shapes the various institutions (gender dimensions of ‘symbolic order’ and ‘institutionalised androcentrism/power of definition’). Also, the filling of the gender dimensions (what they entail) may vary from one policy area to the next. The gender dimensions point to critical societal fields and mechanisms of gendered power relations and rely on empirical evidence that can be generalised for the Global North. They provide an analytical framework that can be applied to other fields and policy areas. A ‘full’ transferability of the GIA approach is, however, not (yet) a given. Fully transferring the GIA approach and educating about its use would require linking insights from these general fields and mechanisms to evidence from the specific policy realms. Further research, based on evidence of sex/gender disparities in the main analytical interests of other policy areas, is needed, which could lead to the refinement of more GIA tools for particular areas of action (not to a ‘one-tool-fits-all’ approach).

The climate GIA was developed with the main implementation context of the UBA in mind, which is designing environmental research projects and giving policy and programme advice to the BMU: The test cases in particular demonstrated that the shape and content of the tool as well as the goal (achieving gender equality and improving climate policy-making) were accepted by users, although research officers and public servants not trained in gender studies or feminist theories did not always find the language of the gender dimensions easy to understand. However, other users – equally unfamiliar with feminist theory – said that for them, the gender dimensions opened their eyes to the relevant issues. The tool is certainly not insubstantial, and engaging with it requires time – as should be allowed for all serious IA procedures. Having said this, after having been educated by the tool, users are asked to decide which gender dimensions they deem relevant for their investigation and purpose. Not all gender dimensions are equally relevant or relevant at the same time. In fact, IA practitioners are encouraged to explore only the most relevant dimensions in-depth, which renders tool application as broad or as narrow as the context and goal of assessment demands.

Undoubtedly, tool implementation depends heavily on institutional commitment and the availability of policy-specific gender expertise (via in-house or external actors). Institutional commitments must also be made to provide information about the tool and integrate it firmly into existing IA procedures.

5. Conclusion

The climate GIA tool discussed here is based on current gender and climate research and operates via the gender dimensions. These aspects render the tool both political and educative and help it serve a communicative function. As there is still a lack of gender knowledge in many fields of climate policy, the tool’s application can enhance the knowledge base on gender issues in these fields. The planning and development of climate policy programmes and measures should be based on the current state of gender-responsive research. The UBA therefore decided to contract this research project in order to base its actions on the results of an international review of the literature on gender and climate and of the analyses of the mechanisms of action. In order to avoid knowledge gaps, a policy-field adapted ex-ante GIA should be applied to all planned climate policies, programmes and projects (and beyond). By employing the new climate GIA tool, the focus shifts from women as a vulnerable group towards the social, institutional and symbolical mechanisms that (re-)produce gender disparities and power imbalances. It also allows for enquiry into the intersectional vulnerabilities of women and men (and where applicable diverse genders). Unlike generic impact assessment tools, the climate GIA has been adapted to the field of climate policies and provides state of the art evidence on sex and gender differences and disparities in this field, helping to educate tool users and support tool acceptance and uptake. In order to avoid being trivialised, it should be carried out under the auspices of gender experts and with strong deliberative elements, including the participation and consultation of women and gender associations committed to climate policy. However, its transformative potential also depends on its accountability framework. Tool application can raise awareness, demonstrate relevance and usefulness and thereby build commitment; on the other hand, it can only be as strong as the commitment of the organisation and the institutional mechanisms the organisation provides, such as employee capacity, expertise, accountability and monitoring mechanisms, and ongoing assessment of how GIA results are actually feeding into policy and programme-making and project design. Sauer (2018) has suggested three E-concepts that summarise the organisational conditions necessary for an effective application of GIA: an organisation must embody the social groups it serves in its staff composition on all levels (men and women equally, while striving for intersectional diversity in staff); embed gender knowledge by, for example, promoting equality issues, hiring gender experts, training employees and so forth; and entrench GIA application and uptake of assessment results firmly in its internal processes and its accountability and quality management structures. Such organisational conditions are the prerequisite to gaining the full epistemic advantage that GIA can provide.

Disclosure Statement

The authors assure that there are no financial interests or benefits linked with this article.

Additional information

Funding

This article is based on the research project ‘Interdependent Gender Aspects of Climate Policy’ (UBA Texts 30/2020) funded by the German Environment Agency (UBA) under grant no. 3716 41 1190 and contracted to the consortium Wuppertal Institute, Institute for Social-Ecological Research and GenderCC. The writing of this article was not part of the project and was conducted without funding/no grants were received.

Notes

1. The climate GIA tool development process and final tool design are published in detail in Spitzner et al. (2020). It represents only one of three work packages; the other work packages were the systematic literature review and a parallel gender analysis of existing German federal climate programmes and policy.

2. One of the few areas next to health impact assessment, in which gender aspects may be closer to the core of assessment is the field of human rights impact assessment (HRIA); see, e.g. the key criteria of the HRIA of the Danish Institute for Human Rights (Götzmann et al. 2016) or for practices and current developments Götzmann (2019).

3. While political and institutional action is required to change IA systems, GIA procedures and oversight structures (Verloo 2008; Lewalter 2013; Sauer 2018).

4. The broad debate around feminist concepts of sex and gender cannot be discussed here in greater detail. Generally speaking, the authors share a social-ecological understanding of gender that draws on different theoretical strands: In accordance with authors such as Sandra Harding, we see gender in GIA as a reflective standpoint (Sauer 2018) and socially constructed category, while the new climate GIA is also informed by recent queer, postcolonial and poststructuralist feminist critique of binaries and homogeneity. At the same time, we share the views of critical feminist scholars on science and technology such as Evelin Fox-Keller, Londa Schiebinger or Donna Haraway, who contend that the physical dimensions of sex form a distinct realm that cannot be reduced to a social construct only (Hummel and Stieß 2017), since corporality is complex and subject to material, social and environmental influences. Strategic essentialism is therefore a guiding principle underlying the climate GIA, and negotiating embodiment against post-structural feminisms.

5. It is noteworthy, that GIA assessments maintain their focus on cis-gendered women and men and cannot replace focused IA on lesbian, gay, bisexual, transgender or intersex issues as i.e. suggested for human rights impact assessment (HRIA) by Sauer and Podhora (2013).

6. Other feminist authors have veered away from gender-focused IA tools altogether and proposed ‘intersectionality assessments’ (Lombardo et al. 2016; Hankivsky et al. 2019). We cannot engage this debate at this point. The primary goals of the climate GIA remains to contribute to gender equality and better climate protection/adaptation through addressing gender dimensions.

7. It is an interesting choice for the editors of a RIA handbook to choose for its single GIA article an author who sees GIA overall in a negative light (too costly, no proof of effectiveness – despite attested lack of application, overtly feminist and therefore politicising) (Gains 2016) – out of the many possible authors and positive voices in the field.

8. The current Covid-19 pandemic shows that policy measures that are only based on epidemiologic evidence fail to account for their wider societal (and economic) impacts and tend to reinforce gender stereotypes and the re-traditionalisation of gender roles in households.