Global assessment of the biodiversity safeguards of development banks that finance infrastructure

Infrastructure development is a major driver of biodiversity loss globally. With upward of US$2.5 trillion in annual investments in infrastructure, the financial sector indirectly drives this biodiversity loss. At the same time, biodiversity safeguards (project‐level biodiversity impact mitigation requirements) of infrastructure financiers can help limit this damage. The coverage and harmonization of biodiversity safeguards are important factors in their effectiveness and therefore warrant scrutiny. It is equally important to examine the extent to which these safeguards align with best‐practice principles for biodiversity impact mitigation outlined in international policies, such as that of the International Union for Conservation of Nature. We assessed the biodiversity safeguards of public development banks and development finance institutions for coverage, harmonization, and alignment with best practice. We used Institute of New Structural Economics and Agence Française de Développement's global database to identify development banks that invest in high‐biodiversity‐footprint infrastructure and have over US$500 million in assets. Of the 155 banks, 42% (n = 65) had biodiversity safeguards. Of the existing safeguards, 86% (56 of 65) were harmonized with International Finance Corporation (IFC) Performance Standard 6 (PS6). The IFC PS6 (and by extension the 56 safeguard policies harmonized with it) had high alignment with international best practice in biodiversity impact mitigation, whereas the remaining 8 exhibited partial alignment, incorporating few principles that clarify the conditions for effective biodiversity offsetting. Given their dual role in setting benchmarks and leveraging private finance, infrastructure financiers in development finance need to adopt best‐practice biodiversity safeguards if the tide of global biodiversity loss is to be stemmed. The IFC PS6, if strengthened, can act as a useful template for other financier safeguards. The high degree of harmonization among safeguards is promising, pointing to a potential for diffusion of practices.


INTRODUCTION
Global biodiversity is declining at unprecedented rates (IPBES, 2019). Infrastructure development is a key driver of biodiversity decline globally (Laurance, 2018). Construction and operation of infrastructure, such as roads, railways, mines, dams, power plants, and ports, drive a range of direct impacts on biodiversity, including through collisions, faunal avoidance behavior, and habitat loss, degradation, and fragmentation (Coffin, 2007). By opening up intact ecosystems, infrastructure development can also induce indirect impacts through increased access and exploitation (Alamgir et al., 2017). Infrastructure development in the transport, extractive, and energy sectors poses a threat to almost half of the species on the International Union for Conservation of Nature (IUCN) Red List of Threatened Species (zu Ermgassen et al., 2019).
The global annual investment in economic infrastructure (i.e., transport, power, telecommunications, and water systems) is rising steadily. It is expected to reach US$2.9 trillion in 2023 and exceed US$3.8 trillion by 2040(Global Infrastructure Hub, 2017. A prominent example of this infrastructure investment boom is China's Belt and Road Initiative (BRI), which is considered the biggest infrastructure development and financing program in history. Announced in 2013, BRI encompasses extensive transboundary transport infrastructure networks flanked by thousands of energy, mining, and industrial projects (Teo et al., 2019). Many of these networks are being built along 7 corridors (6 overland and 1 maritime) connecting key cities and ports across 70 countries in Asia, Europe, and Africa and the adjacent seas (Ruta et al., 2019). According to a World Bank report, the total lending to BRI projects is already US$575 billion (Ruta et al., 2019). The G7 (7 of the world's largest and wealthiest economies) have also recently announced the Build Back Better World initiative, a Western-led counter to the BRI, which is expected to mobilize billions more in infrastructure investment (Reuters, 2021).
Traditionally, infrastructure was almost exclusively funded by public financiers (including governments and development banks). Over time, the availability of public funds started declining due to rising deficits and budgetary constraints. In the 1990s, private banks started making a foray into the infrastructure space. By the mid-2000s, public and private institutional investors, such as pension funds, sovereign wealth funds, mutual funds, and insurance companies, also started investing in infrastructure projects (OECD, 2014). Typical large infrastructure projects of today are financed by a combination of all these actors (Inderst & Stewart, 2014;Missbach, 2004;OECD, 2015) ( Figure 1).
Although the figures vary widely with country, public-sector contribution has been estimated to account for about threequarters of the total infrastructure financing in developing countries, with the private sector accounting for the remaining. In the public sector, 56% of the total infrastructure spending comes from government budgets, and development finance accounts for about 18% of the total (Bhattacharya et al., 2012;OECD, 2015) (Figure 2). Although its contribution has hitherto been limited, the private sector (banks and institutional investors) is increasingly being seen as an important source of infrastructure financing (OECD, 2014). Likewise, national development banks (around 250 in number and holding almost US$5 trillion in assets) are poised to become key players in infrastructure financing not just in their own countries, but also internationally (Studart & Gallagher, 2016). A few national development banks such as China Development Bank are already internationalizing, having extensively financed infrastructure projects in developing countries (Horn et al., 2019).
New infrastructure development is increasingly coinciding with areas of high biodiversity value (Laurance et al., 2014;Rehbein et al., 2020;Zarfl et al., 2015). By choosing to invest in infrastructure projects located in such areas, financiers indirectly drive biodiversity loss. At the same time, biodiversity safeguards required by financiers can help limit some of this damage (Gallagher & Yuan, 2017;Morgado & Taşkın, 2019). Biodiversity safeguards are impact mitigation measures that financiers require developers to apply to a project as a condition of funding (Larsen & Ballesteros, 2013).
The coverage and harmonization of biodiversity safeguards are important factors in their effectiveness and therefore warrant scrutiny (Gallagher & Yuan, 2017;Himberg, 2015). It is equally important to examine the extent to which these safeguards represent international best practice. We focused our analyses on the biodiversity safeguards of infrastructure financiers in the development finance space (i.e., public development banks and nonbank development finance institutions [hereafter, development banks]). These are international and FIGURE 2 Infrastructure financing by financier type (Bhattacharya et al., 2012;OECD, 2015). national public-sector financial institutions with an exclusive policy mandate of funding development projects (Xu et al., 2020) (i.e., projects intended to spur economic or social development). Even though development banks contribute only about 18% to global infrastructure finance (Bhattacharya et al., 2012;OECD, 2015), the reason for a focus on them is 3-fold. First, multilateral development banks (MDBs) (i.e., development banks with multilateral ownership), such as the World Bank Group, are the pioneers of environmental and social safeguard systems (Caspary, 2009). Second, MDBs are seen as catalysts for leveraging private finance into the infrastructure space. They often cofinance infrastructure projects with other financial institutions (such as private banks and institutional investors) in order to absorb the intrinsic risks associated with the capital-intensive and long-term nature of such projects. In the process, they can act as agents for transfer of good practice to other cofinanciers (Chelsky et al., 2013;Gallagher & Yuan, 2017). Third, when projects are cofinanced by several financial institutions, the most stringent set of safeguards are usually applied, which, in most cases, are those of development banks (ITUC, 2019). Although MDB safeguards are widely recognized (Bennun et al., 2018), little is known about the coverage of biodiversity safeguards among other development banks.
In addition to coverage, harmonization among financier safeguards is also an important factor in effective impact mitigation because it precludes shopping around by borrowers for less onerous requirements (Gallagher & Yuan, 2017). There has been a conscious effort to achieve harmonization across international financial institutions; several MDBs have benchmarked their safeguards against International Finance Corporation (IFC) Performance Standards (Gallagher & Yuan, 2017;Horberry, 2014). Compliance with IFC Performance Standards is also built into the Equator Principles, a voluntary safeguard system that has been adopted by over 100 international and national private financial institutions in 37 countries (Equator Principles, 2022), and into the "common approaches," a set of mandatory standards adopted by Organization for Economic Co-operation and Development (OECD) country export credit agencies (Gallagher & Yuan, 2017). The Association of Bilateral European Development Financial Institutions (EDFI) also requires its 15 members to adhere to IFC Performance Standards (EDFI, 2017). Despite these efforts, inconsistencies remain among the safeguards of development banks (Caspary, 2009;Gallagher & Yuan, 2017), which means an examination of the extent to which harmonization has taken place is timely.
Biodiversity safeguards of financiers are often structured around the mitigation hierarchy, a framework that allows for successively avoiding, minimizing, remediating, and offsetting impacts, often with the overall goal of at least a no net loss of biodiversity (Himberg, 2015;IFC, 2012). A number of best-practice principles fundamental to the successful application of the mitigation hierarchy have been outlined in 2 multistakeholder-consensus-driven policies: the IUCN Policy on Biodiversity Offsets (IUCN, 2016) and the BBOP (Business and Biodiversity Offsets Programme) Standard on Biodiversity Offsets (BBOP, 2012b). The IUCN policy was arrived at through an expert-led examination of the scientific literature to identify the ecological principles that underpin the best practice in biodiversity impact mitigation, an evidence-based analysis of existing offset schemes to identify practical issues in the implementation, and consideration of national and regional context (IUCN, 2016). Similarly, the BBOP Standard on Biodiversity Offsets was developed after extensive consultations among its vast membership, spanning industry, governments and intergovernmental organizations, conservation groups, and academia (BBOP, 2012b(BBOP, , 2018. Both policies are routinely considered international best practice (De Witt et al., 2019;Sullivan, 2013). The extent to which biodiversity safeguards of development banks align with international best practice is an important indicator of their effectiveness in impact mitigation.
We examined the coverage of biodiversity safeguards among development banks and how this varied with type of ownership and geographic scope of operations. The biodiversity safeguards we found were assessed for alignment with best-practice principles of biodiversity impact mitigation outlined in the IUCN and BBOP policies. We also examined the extent to which these safeguards were harmonized across development banks. We sought to identify broad patterns in coverage, harmonization, and alignment with best practice, with a view toward offering insight into the current state of biodiversity safeguards of development banks.

Selecting development banks
To identify the development banks for our assessment, we used the Global Database of public development banks and development finance institutions compiled by Peking University's Institute of New Structural Economics and Agence Française de Développement (https://www.nse.pku.edu.cn/ dfidatabase/), which had 527 records as of October 2021 (INSE & AFD, 2022). The database lists development banks as per their official mandate and assets under management. We selected banks with official mandates that align with highbiodiversity-footprint infrastructure investment: infrastructure, international financing of private sector development, flexible, and export or import. Banks with mandates of investing in agriculture, social housing, local governments, and micro, small, or medium enterprise were excluded because these are unlikely to invest in high-biodiversity-footprint infrastructure. To make the assessment feasible within a reasonable time frame, we also excluded the banks classified as micro in size (<US$500 million in assets) because these are also unlikely to have safeguards. Following these protocols, 155 development banks became the focus of our assessment.

Assessing coverage of biodiversity safeguards and extent of harmonization
Before examining the presence of biodiversity safeguards specifically, we assessed whether each of the 155 shortlisted development banks had any environmental safeguard systems (i.e., explicit requirements for environmental impact assessment and management) in place. For this, we carried out internet searches with the following search string: <bank name> AND (environment* OR green OR ecolog*) AND (polic* OR safeguard* OR standard* OR framework OR guideline* OR strateg* OR plan*). We also visited the websites of each of the banks to check for information on environmental safeguards. For banks with websites in a language other than English, we carried out keyword searches in the respective language. The presence or absence of environmental safeguard systems was recorded for each bank. An important caveat here is that because we used internet search as a method to record the presence or absence of safeguards, a bank recorded as having safeguards as absent could still have safeguards as a part of their internal requirements even if the safeguards were not discoverable online.
The identified systems of environmental safeguards were then examined for presence of requirements or measures specific to biodiversity impact mitigation. The entire text of each bank's safeguard documentation was read with particular attention to terms such as biodiversity, conservation, ecosystems, ecological, flora and fauna, nature or natural environment, species, and protected areas.
We also examined how the coverage of safeguards varied across different types of development banks. The Institute of New Structural Economics and Agence Française de Développement's database provides information on ownership of development banks (multilateral, national, subnational) and on the geographical scale of their operations (international, national, regional, or subnational). We combined these 2 parameters to arrive at 5 types of development banks consistent with the nomenclature of OECD and Overseas Development Institute (ODI) (Engen & Prizzon, 2018) (Table 1; Appendix S1). We compared the presence of biodiversity safeguards across these 5 types of development banks. We also examined the distribution of development banks with biodiversity safeguards across continents.
In addition to coverage, we assessed the extent of harmonization among the identified safeguards. For this, documents were reviewed for the presence of any text that indicated that the safeguards were benchmarked against IFC Performance Standards (or the safeguards of another MDB). All of the OECD country export credit agencies in our study were recorded as containing biodiversity impact mitigation measures and aligning with IFC Performance Standard 6 (PS6), given that member countries have agreed to adhere to OECD's "common approaches," which in turn are benchmarked against IFC Performance Standards. Likewise, banks that were signatories to the Equator Principles or members of EDFI were also recorded as having biodiversity impact mitigation measures and aligning with IFC PS6.

Assessing alignment with international best-practice principles
We used the IUCN Policy on Biodiversity Offsets (IUCN, 2016) and the BBOP Standard on Biodiversity Offsets (BBOP, 2012b) to identify best-practice principles of biodiversity impact mitigation. These are conditions that the policies say must be met for appropriate application of the mitigation hierarchy. To identify the principles, the entire text of the policy documents was read; special focus was given to statements with the words must or should. Some of the principles were pertinent to social and cultural impact (e.g., IUCN policy has a principle that requires following a rights-based approach). Such principles, although important, were not a focus of this study and were excluded. There is considerable overlap between the principles listed by both the policies. We included biodiversity impact-focused principles that featured in either or both of the policies (Table 2; Appendix S3).
Next, we carried out content analysis of the development banks' safeguard documents to ascertain whether the biodiversity impact mitigation measures specified in them aligned with the identified best-practice principles. We recorded whether the document made a strong reference, weak reference, or no reference to each principle. The text indicating the reference was also recorded. A strong reference was recorded when the principle was referred to in its entirety and a weak reference was recorded when only 1 aspect of the principle was referred to. To illustrate, a safeguard document was recorded as making  a strong reference to the principle of full range of impacts, which entails the assessment of direct, indirect, and cumulative impacts, if all 3 types of impacts were mentioned and a weak reference if only 1 or 2 types were mentioned.

Coverage and consistency of biodiversity safeguards
Environmental safeguard systems were detected online for <50% (77 of 155) of the 155 development banks assessed (Appendix S1). Biodiversity impact mitigation measures were detected for <42% (65 of 155) of the development banks (Figure 3; Appendix S2). A greater proportion of banks with international operations (irrespective of ownership) (i.e., MDBs, bilateral development agencies, export credit agencies, and national banks with international operations) had biodiversity safeguards compared with banks with only national operations (Figure 4; Appendix S2). A greater proportion of European development banks had biodiversity safeguards compared with those in other continents ( Figure 5; Appendix S2). More than 86% of the development banks with biodiversity safeguards (56 of 65) were identified as explicitly benchmarked against the IFC Performance Standards (Figure 3; Appendix S2). The remaining 8 were divergent (i.e., had their own safeguard systems not benchmarked against IFC Performance Standards). These included Asian Development Bank, African Development Bank, Asian Infrastructure Investment Bank, New Development Bank, West African Development Bank, Caribbean Development Bank, Development Bank of Southern Africa, and Infrastructure Development Company Limited, Bangladesh (Appendix S2).

Alignment with best-practice principles of biodiversity impact mitigation
We identified 20 best-practice principles from the IUCN and BBOP policies (Table 2) (Appendix S3 contains a description of each principle and how it featured in the IUCN and BBOP policies). The principles included requiring strict adherence to and the early application of the mitigation hierarchy, accounting for the full range of impacts, exploring all project alternatives, adoption of the precautionary, adaptive management and integrated planning approaches, and monitoring and evaluation of mitigation measures. Some principles were specific to the offsetting stage of the mitigation hierarchy; these included opting for offsets only as a last resort, not using offsets to justify the project, recognizing limits to and risk of failure of offsetting, designing offsets to achieve measurable outcomes such as no net loss or net gain in line with international or national conservation goals, and using appropriate metrics and counterfactuals to assess the offset outcomes. Principles requiring additionality, permanence, and ecological equivalence of offsets were also identified. The IFC PS6 (and by extension the safeguards of the 56 development banks that are benchmarked against IFC) had the highest level of alignment with best-practice principles; 19 out of 20 principles were mentioned. The only principle that was not mentioned in IFC PS6 was the principle of prevention of leakage, which precludes offsetting by simply displacing biodiversity pressures (such as deforestation, poaching, etc.) to another location instead of addressing the problem. The policies of the 8 divergent banks that included prominent MDBs, such as African Development Bank, Asian Development Bank, Asian Infrastructure Investment Bank, and New Development Bank, aligned with fewer than half of the identified principles. Principles clarifying the conditions for effective biodiversity offsetting were particularly absent (Table 2; Appendix S3).
Several key best-practice principles were represented widely in bank safeguards. For example, the project alternatives principle (which requires project developers to demonstrate that no other viable lower impact alternatives, in terms of location and design, to the proposed project exist) was mentioned in the policies of all 65 banks that had biodiversity safeguards. Similarly, biodiversity safeguards of all the banks made either a strong or weak reference to the full-range-of-impacts principle that requires developers to take full account of direct, indirect, and cumulative impacts of the project. Biodiversity safeguards of all banks, except Caribbean Development Bank, made a reference to the adherence-to-the-mitigation-hierarchy principle, which requires following each step of the mitigation hierarchy sequentially and using offsets only as a last resort. Likewise, all banks mentioned the monitoring-and-evaluation principle, which requires project developers to put in place mechanisms for rigorous and long-term monitoring and evaluation of biodiversity impact mitigation measures applied during the project. All the banks, except the African Development Bank, required project developers to adopt a precautionary approach (i.e., apply mitigation measures even when there is uncertainty on the scientific evidence on the impacts) ( Table 2; Appendix S3).
Principles that were poorly represented in safeguards other than IFC PS6 included those relating to the use of effective metrics and counterfactuals to measure no net loss or net gain and alignment of these mitigation goals with national or international conservation goals. Similarly, principles that preclude offsetting when perverse outcomes, such as cost shifting, might result; when vulnerable or irreplaceable biodiversity is involved; or when the time lag between project impact and offset gains is inordinate were poorly represented. Similarly, the principle requiring offsets to be maintained in perpetuity was infrequently mentioned (Table 2; Appendix S3).

Coverage and coherence of biodiversity safeguards
Our assessment of the coverage of biodiversity safeguards across development banks investing in infrastructure revealed significant gaps. Biodiversity safeguards were detected online only for 42% of the 155 assessed development banks. Our results corroborated the findings of a 2020 study that concluded that China Development Bank and China Export-Import Bank (2 of the top 3 development banks in terms of assets) had no biodiversity safeguards (Narain et al., 2020).
It is evident that development finance, which constitutes about 18% of the global infrastructure financing, is not subject to adequate impact mitigation measures when it comes to biodiversity. This is despite the fact that development finance routed without effective safeguards causes damage to nature potentially worth US$800 billion every year, especially in resource-intensive countries with high levels of biodiversity and relatively weak regulation (Finance for Biodiversity Initiative, 2021). Our assessment of coverage of safeguards was limited to the shortlist of 155 development banks (out of the 500+ in the global database). Banks with <US$500 million in assets or with mandates that do not align with high-biodiversity-footprint infrastructure investment were excluded because of the lower likelihood of them having biodiversity safeguards. However, there is still a small chance that some of the excluded banks have biodiversity safeguards in place.
Although our findings cannot be directly extrapolated to the entire gamut of infrastructure financiers (which also include private banks and national governments), the fact that very few international safeguard standards exist outside the development finance space (e.g., Equator Principles) points to much lower coverage of safeguards across the breadth of infrastructure financing. Many jurisdictions have their own legal and regulatory systems for environmental impact assessment and management, which can provide a layer of protection for biodiversity. However, many country systems have been found by the World Bank to be below par compared with international standards (Priester, 2019). In fact, a recent global assessment showed that half of the countries with planned infrastructure expansion lack mandatory policies on management of biodiversity impact (zu Ermgassen et al., 2019). This makes the limited coverage of safeguards among infrastructure financiers even more concerning.
What is potentially promising is the high degree of harmonization among the safeguards; 56 (86%) of them were benchmarked against IFC PS6. In many cases, regional development banks adopt safeguards that are best suited to their own development context and choose not to harmonize with the safeguards of global development banks (Mbengue & de Moerloose, 2017). However, where harmonization does occur, it provides an opportunity for transfer of best practice (Gallagher & Yuan, 2017).

Alignment with best practice
Our results also showed that IFC PS6 (and by extension the safeguards of the 56 banks that are benchmarked against IFC) comes close to international best practice in biodiversity impact mitigation. This is promising because IFC alone invests about US$1.5 billion in infrastructure projects every year (IFC, 2020). However, we also found that safeguards of the 8 divergent banks (which include 4 MDBs, namely, Asian Development Bank, African Development Bank, Asian Infrastructure Investment Bank, and New Development Bank) only partially align with international best practice. Particularly, principles pertaining to the effective design, implementation, and governance of offsetting were almost entirely absent from the safeguards of the divergent banks even as they aligned with the principle of adherence to the mitigation hierarchy as a whole. Offsetting is an integral part of the mitigation hierarchy and clarifying the conditions for effective offsetting is important for its application. For example, none of the divergent banks subscribed to the limits-to-offsets principle that requires project developers to recognize that certain impacts cannot be offset and the project should not proceed if such impacts are likely to occur. It has been established that impacts on irreplaceable and highly vulnerable elements of biodiversity cannot be offset (BBOP, 2012a;Phalan et al., 2018). A recent study, for instance, showed that the impacts of deep-sea mining cannot be adequately offset due to the highly vulnerable nature of deepsea biodiversity and pointed to avoidance as the only recourse (Niner et al., 2018). It is therefore important for financiers to clearly specify the requirement of limits to offsets in their safeguards.
A few of the divergent banks, including the 2 new non-Western-backed MDBs, Asian Infrastructure Investment Bank and New Development Bank, failed even to include the principle of measurable conservation outcomes that requires specifying the mitigation goals of no net loss or net gain. Specific conservation goals provide project developers a clear direction for applying mitigation measures (Simmonds et al., 2020). It is important that the divergent banks make an effort to incorporate these best-practice principles in their safeguards more comprehensively.
The IFC PS6 needs to be strengthened to act as a better benchmark for other development banks, as well as private and institutional investors. Several key principles that were absent from the safeguards of the divergent banks were also only cursorily referred to in IFC PS6. For example, the principle that requires project developers to specify the reference counterfactual for the required goals (no net loss or net gain) was referred to in IFC PS6 only in passing. A counterfactual scenario is a scenario that would have happened in the absence of the offset. It is important for safeguard policies to require project developers to make explicit and transparent assessments of the plausibility of the assumed counterfactuals because the very nature of a counterfactual (a future that will never be observed) makes it prone to manipulation (Maron et al., 2016). Overstating a counterfactual of biodiversity decline, for example, will falsely yield smaller offsetting requirements, less biodiversity gains from offsetting, and a net loss in biodiversity across the jurisdiction Watson et al., 2010).
Another principle, alignment with national and international conservation goals, was similarly absent across the divergent bank safeguards and only implied in IFC PS6. This principle requires designing offsets in such a way that they align with, or make a proportional contribution toward, jurisdiction-level goals (Simmonds et al., 2020). For example, the amount and type of offset for a given loss can be determined such that the offset gain helps achieve goals, such as recovering threatened species or ensuring that ecosystem extent and condition do not decline, in line with the jurisdictional goal. Global agreements like the Convention on Biological Diversity provide a useful framework for determining such jurisdiction-level goals (CBD, 2022), and financier safeguards can and should embed principles that ensure offsetting contributes to key jurisdictional (and global) biodiversity imperatives.
The principles that preclude offsetting when perverse outcomes, such as cost shifting or leakage, might result were similarly either not mentioned or only cursorily mentioned even in IFC PS6. Cost shifting refers to use of offsets to fund existing conservation commitments in violation of the principle of additionality . Leakage can occur when the offset merely relocates the pressures of biodiversity loss (e.g., deforestation or poaching) to another site instead of stemming them (Moilanen & Laitila, 2016). Both cost shifting and leakage defeat the purpose of the offsetting and lead to net loss and therefore should be explicitly cautioned against in safeguard requirements. Likewise, the principle of permanence, which requires offset sites to be maintained at least as long as the impact lasts and preferably in perpetuity, was absent from the safeguards of the divergent banks and had only a weak mention in IFC PS6. Project impacts are often permanent, which requires that conservation gains resulting from the offset are also permanent (Moilanen & Laitila, 2016). It is important that project developers ensure permanence through appropriate legal mechanisms (Virah-Sawmy et al., 2014).
Although alignment with best-practice principles is an indicator of the effectiveness of safeguards, it does not automatically mean that on-the-ground biodiversity outcomes are positive (the latter is also contingent upon other factors, including interpretation of the requirements, due diligence, and supervision by bank staff) (Richard, 2017). The limited available assessments of the actual biodiversity outcomes of development bank projects show mixed results and necessitate more extensive examination of the impact of biodiversity safeguards on the actual biodiversity outcomes of projects. A recent study of World Bank projects funded from 1995 to 2014, for example, shows that although project siting successfully avoided protected areas, there remained a high spatial coincidence with other important areas for biodiversity, including key biodiversity areas and ranges of globally threatened birds, mammals, and amphibians (Morley et al., 2021). Another study showed that the rates of tree loss in important bird areas located adjacent to World Bank projects approved from 2000 to 2011 were not higher than that in the important bird areas located at a distance from the projects (Buchanan et al., 2018). Thus, there is limited evidence to show the extent to which World Bank safeguards (which have requirements of net gain in critical habitat on the lines of IFC PS6) have managed to achieve their stated goals.

Way forward
Infrastructure development fuels economic growth; it facilitates extraction of resources, generation of power, manufacturing and trade of goods, and movement of people (Doyle & Havlick 2009). Investment in basic infrastructure (meant for provision of basic services, such as water, sanitation, and electricity) is seen as paramount for attainment of several of the United Nation's Sustainable Development Goals (Thacker et al., 2019), a set of globally agreed aspirations that will determine the world's development agenda over the next decade (UN Desa, 2016).
Infrastructure development, however, often comes at a cost to biodiversity. Poorly sited, designed, and executed infrastructure projects chip away at fragile ecosystems (Laurance et al., 2015), generating externalities that disrupt the flow of ecosystem services (benefits that biodiversity provides to human beings) and diminishing the very development dividends they are meant to provide (Mandle et al., 2016).
The UN Convention on Biological Diversity's Post-2020 Global Biodiversity Framework, which was adopted by the parties in December 2022 (IUCN, 2022), is slated to reset the world's conservation agenda through 2030. Many of the action targets outlined in the framework are centered around tackling the drivers of biodiversity impact (CBD, 2022). The sheer quantum of funds that flow into infrastructure globally shows that any amount of international ambition to protect biodiversity will at best be reactive unless drivers such as infrastructure financing are more proactively addressed. To effectively mitigate the biodiversity impact that they drive, infrastructure financiers need robust and comprehensive biodiversity safeguards that align with international best practice. This is imperative if the goal of stabilizing biodiversity loss by 2030 envisaged in the Post-2020 Global Biodiversity Framework is to be achieved.
Our results, however, revealed a low coverage of biodiversity safeguards among development banks that fund infrastructure. Our results also showed that biodiversity safeguards were concentrated among development banks with international operations. But the high degree of harmonization among the existing safeguards is a good sign, pointing to a potential for their diffusion among remaining financial institutions. Also, promisingly, in 2020, 450 development banks from around the world signed a joint declaration committing to cooperate among themselves and with private actors, to align all financial flows with the Post-2020 Global Biodiversity Framework (Finance in Common, 2020). Going forward, IFC PS6, if strengthened, can act as a benchmark not only for private banks and institutional investors, but also for national development banks, whose role in infrastructure financing is slated to grow.