Fragmentation and Coordination of REDD+ Finance under the Paris Agreement Regime

: Under the Kyoto Protocol regime, various forms of ﬁnancial support have been committed to helping the implementation of reducing emissions from deforestation and forest degradation, as well as fostering conservation, the sustainable management of forests, and the enhancement of forest carbon stocks (REDD+) in developing countries. We analyzed the fragmentation of REDD+ ﬁnance and suggested methods for its coordination under the Paris Agreement regime. The fragmentation of REDD+ ﬁnance was observed, but it was lower than that of general ofﬁcial development assistance (ODA). However, we found that the trend of fragmentation in REDD+ ﬁnancing is different from that of general ODA, with a few major donors occupying a large portion of the total size of committed REDD+ ﬁnance. Thus, it may not be appropriate to consider the fragmentation of REDD+ ﬁnance merely as an obstacle that needs to be decreased. Still, the total amount of REDD+ ﬁnance should be increased and adjusted for various donor–recipient relationships, in consideration of the REDD+ ﬁnance options in the Paris Agreement. Some REDD+ countries have made progress in national REDD+ and accomplished emission reductions. However, REDD+ ﬁnance needs to be stratiﬁed considering the progress of national REDD+. For such forms of cooperation, an information-sharing and monitoring system that collects information on ongoing REDD+ cooperation, the commitment and expenditure of REDD+ ﬁnance, and the support needs of REDD+ countries at a global level should be established. Multilateral organizations need to provide safeguarding functions for developing countries that are isolated from bilateral REDD+ ﬁnance.


Introduction
The global climate regime has transitioned from the Kyoto Protocol to the Paris Agreement from 2021, and conserving and enhancing sinks and reservoirs of greenhouse gases are undoubtably important. To aid in reducing emissions from deforestation and forest degradation and fostering conservation, the sustainable management of forests, and the enhancement of forest carbon stocks (REDD+), an incentive mechanism for reducing greenhouse gas (GHG) emissions in the forest sector was developed through negotiations and decisions under the Kyoto Protocol regime. Under the Kyoto Protocol regime, developing countries-the non-Annex I countries of the Kyoto Protocol-were encouraged to undertake forest sector mitigation actions and receive result-based finance [1]. Article 5 of the Paris Agreement acknowledges the existing framework of REDD+ and encourages its implementation and support. Thus, under the Paris Agreement, REDD+ will be important for finance. Based on our analysis, we provide several recommendations for the coordination of fragmented REDD+ finance in order to contribute to more effective mitigation through REDD+ under the Paris Agreement regime. To this end, we propose the following research question: How fragmented is the REDD+ finance conducted under the Kyoto Protocol regime? The remainder of this study proceeds as follows. In Section 2, we discuss the characteristics of REDD+ finance and its fragmentation in detail. In Section 3, the empirical data and methods are described. In Section 4, the results of our analysis are presented. In Section 5, our results are discussed. Section 6 concludes this study.

Transition of the Climate Regime and Characteristics of REDD+ Finance
As subsidiary agreements of UNFCCC, the Kyoto Protocol and Paris Agreement form the foundations of the Kyoto Protocol and Paris Agreement regimes. Although they have a common goal to maintain the global average temperature below 2°C above the pre-industrial level, differences exist in several aspects; for example, the scope of the parties responsible for GHG mitigation and approaches in which mitigation activities are approved. Accordingly, characteristics of REDD+ finance under the Paris Agreement and Kyoto Protocol regimes have both similarities and differences.
In the Kyoto Protocol regime, only developed countries belonging to Annex I were responsible for mitigation. Though developing countries did not have such a responsibility, REDD+ was proposed and developed in UNFCCC negotiations to encourage developing countries to reduce emissions in the forest sector [20]. This emerged in 2005 as Reducing Emissions from Deforestation (RED) and was quickly expanded to REDD+ through consecutive negotiations; the framework of REDD+ implementation, WRF, was adopted at COP 19. Considering respective capabilities and national circumstances and different levels of support received, developing countries may take a phased approach-Readiness (phase 1), Implementation (phase 2), and RBP (phase 3)-to implement REDD+ (Table 1). With these institutional environments, REDD+ finance serves as an aid to developing countries to support their REDD+ implementation capacity and receive compensation without any returns to the "donor" developed countries or institutions that provided REDD+ finance. Table 1. REDD+ phased approaches [3]. Under the Paris Agreement, in contrast to the Kyoto Protocol, all 196 parties, which also include developing countries, are responsible for climate action: to this end, parties are required to set, regularly update, and strive to achieve the NDCs, which are the national mitigation and adaptation objectives of each party [21]. Article 6 of the Paris Agreement allows parties to transfer ITMOs between themselves through cooperative approaches and utilize ITMOs to achieve NDCs, while the relevant specific guidelines are not agreed yet; Article 5 of the Paris Agreement acknowledges the existing REDD+ frameworks and encourages parties to take actions to support and implement REDD+ [22]. Accordingly, it is highly likely that REDD+ will be used as a means not only to achieve a host country's NDC but also to secure and use ITMOs for the partner countries' NDCs (Table 2).

Donors' Motivations for REDD+ Finance
Considering that REDD+ finance has similarities with aid, studies on donors' motivation for aid allocation can provide insights on the motivations of donors in REDD+ finance. These studies suggest that recipient needs for aid [23,24] and recipient merits that may impact effectiveness and efficiency of aid [25][26][27] are some of the determinants of aid allocation; donor interest was also found to be an important determinant, as aid is a part of the donors' foreign policy, through which donors seek political and economic benefits [26,[28][29][30]. It was also found that, under the Kyoto Protocol regime, donors of REDD+ finance have considered recipient needs, recipient merits, and donor interests when allocating finance to REDD+ countries [31].

Fragmentation of REDD+ Finance
Aid fragmentation, which refers to "aid to a recipient country received from too many sources and spread over too many programs" [32], becomes severe as the number of donor countries and organizations and recipient countries increases. Aid fragmentation has been reported to negatively impact the effectiveness of aid by causing problems such as an increase in transaction costs, a decrease in the amount of aid allocated to individual recipient countries, and an increase in the administrative demands in recipient countries [33,34]. Under the Kyoto Protocol regime, the size of REDD+ finance and the number of donor countries, organizations, and recipient developing countries have increased rapidly in line with the institutionalization of REDD+ [19], which has deepened the fragmentation of REDD+ finance. It is expected that the fragmentation of REDD+ finance will become serious under the Paris Agreement regime, as all parties are obliged to make efforts for mitigation; actions for REDD+ implementation are also encouraged. More actors, including both donor and recipient countries and even some organizations, will attempt to use REDD+ to meet their obligations. For example, 56 countries mentioned in their NDCs explicitly that they were considering REDD+ as a means to achieve their national emissions reduction targets [35].
Considering the financing motivation of donors and relevant decisions of UNFCCC, REDD+ financing under the Paris Agreement regime will serve an important role in supporting developing countries to establish elements of WRF, delivering the RBPs to developing countries for their REDD+ activities, as well as implementing cooperative approaches and securing ITMOs.

Data Collection
We collected data on bilateral and multilateral REDD+ finance commitments from 2005 to 2019 to analyze the distribution and fragmentation of REDD+ finance. Among the REDD+ finance cases reported in the Food and Agricultural Organizations Voluntary REDD+ Database (FAO VRD) [36] and Overseas Development Institute Climate Funds Update (ODI CFU) [37], 25 donors (16 countries and nine organizations) and 108 recipient countries were selected for analysis. For the analysis, 2110 cases from FAO VRD and 384 cases from ODI CFU were analyzed.

Fragmentation of REDD+ Finance
We applied the fragmentation and concentration rate estimation method presented in [32], which considers both the perspectives of donors and recipient countries, to analyze the degree of fragmentation of REDD+ finance. Assuming that there is a donor-recipient relationship in which the recipient country receives a much smaller amount of REDD+ finance compared to the corresponding donor's other recipients, the importance of this relationship for the donor may not be significant. However, if this donor is a major or the only donor for the recipient country, the importance of this relationship for the recipient country will be significant. In this context, if the significance of committed REDD+ finance is low for both a donor and a recipient tied in a donor-recipient relationship, such finance is considered not to be significant.
Based on this proposition, we conducted two analyses on the donor-recipient tie. First, we analyzed whether a donor supported a specific recipient country in a higher proportion than the ratio of its total amount of REDD+ finance to the total global amount of REDD+ finance. Assuming donor A provides REDD+ finance to recipient B, recipient B is considered significant from donor A's perspective if Size of finance from donor A to recipient B Total size of donor s REDD + finance ≥ Total size of donor s REDD + finance Total size of global REDD + finance (1) Otherwise, the recipient is not significant. Second, from the perspective of the recipient country, we ranked all the donors that provided REDD+ finance to this recipient by the respective size of REDD+ finance and analyzed whether a specific donor belonged to the top 90%. Let D = {1, 2, ..., n} be a set of n donors who provide REDD+ finance to a specific recipient country and F = {f 1 , f 2 , ..., f n } be a set of the amount of REDD+ finance commitments of each donor in n; then, donor j in n is significant from the perspective of this specific recipient country if where ∑ n i=1 f i refers to the sum of committed REDD+ finance of all n donors to this specific recipient country, and f j is the amount of REDD+ finance commitment of the jth donor and an element of F.
For the significant/not significant relationships, respectively, from the perspective of a donor and a recipient, four categories of donor-recipient relationships in the global REDD+ finance system can be identified: donor-recipient relationships that are significant both for donors and recipient countries (group A); donor-recipient relationships significant only for donors (group B); donor-recipient relationships significant only for recipients (group C); donor-recipient relationships that are not significant for either donors or recipients (group D).
Based on this categorization, we calculated the ratio of fragmentation at the actor level and the global level, respectively, from the donors' and recipients' sides. Following [32], it can be assumed that fragmentation is higher if the share of insignificant relationships to total relationships is higher. Therefore, the "fragmentation rate" of a country is calculated as (D/(A + B + C + D)); the fragmentation rate of the global REDD+ finance system is calculated as the mean of the fragmentation rate of all countries. The potential to improve the efficiency in REDD+ finance is high for the donors and recipients that have a higher fragmentation rate than that of the global REDD+ finance system.
Besides, from the definition of aid fragmentation, it can be assumed that, other conditions being equal, the degree of fragmentation has (1) a positive correlation with the total number of donors and (2) a negative correlation with the total size of commitments. Thus, we investigated whether such relationships were observed in the case of global REDD+ finance by investigating the fragmentation of each recipient REDD+ country.

Classification of REDD+ Countries by REDD+ Progress
To suggest options for the adjustment of fragmented REDD+ finance under the Paris Agreement regime, we grouped REDD+ countries by their REDD+ implementation progress, referring to the information submitted to the UNFCCC REDD+ Web platform. If a REDD+ country has more than one registered case of ERs on the platform and/or received RBP results, it is considered to have fulfilled the WRF elements, as well as to be in phase 3 and classified as group I; if a REDD+ country listed on the website does not have registered ER but has submitted FREL/FRL and received technical assessments, it is considered to be in phase 2 and classified as Group II; if a REDD+ country has registered a national focal point or is listed on the platform (without any further information), it is considered to be in phase 1 and classified in Group III; if a country has not submitted any documents relevant to WRF on the platform, it is considered to be not engaged in WRF and classified as group IV. For each group, the number of recipient countries and donors, allocated amounts of REDD+ finance, and fragmentation rate were analyzed to suggest the coordinating options under the Paris Agreement regime. Table 3 shows the 457 bilateral and multilateral REDD+ finance cases between 16 donor countries, 9 donor organizations, and 108 recipient countries that were reported to the FAO VRD and ODI CFU between 2006 and 2014. The total amount of commitment was 9699.27 million USD; of which 63.9% (6197.68 million USD) was bilateral and 36.10% (3501.59 million USD) was multilateral, showing that the amount of bilateral commitment is larger than that of multilateral commitment.  Table 4 shows the fragmentation rate by country groups with major forest types and global REDD+ finance. The number of recipient countries by major forest type was the largest in the tropical country group (87), followed by the sub-tropical group (13), and the temperate and boreal group (8). The number of donor-recipient relationships by continent was the largest in the tropical groups (412), followed by sub-tropical (29), and temperate and boreal (16). The received amount was the largest in the tropical country group (8881.21 million USD), followed by the sub-tropical group (90.25 million USD), and the temperate and boreal group (390.25 million USD). The fragmentation rate was the highest in the tropical group (22.57%), followed by the sub-tropical group (17.24%), and the temperate and boreal group (12.50%).  Figure 1 shows the distribution of recipient REDD+ countries in which the x-axis is the number of donors and the y-axis is the fragmentation rate. Figure 2 displays the relationship between the fragmentation rate (y-axis) and the natural logarithm of the committed amount of finance (x-axis) among the recipient REDD+ countries. The statistics regarding the amounts of commitments, the number of donors, and fragmentation rate of each recipient country is summarized in Table 5. In both figures, it is assumed that the recipient countries whose fragmentation rate is higher than that of the total global REDD+ finance (21.88%) have a high degree of fragmentation.

Overview of REDD+ Finance
global REDD+ finance. The number of recipient countries by major forest type was th largest in the tropical country group (87), followed by the sub-tropical group (13), and th temperate and boreal group (8). The number of donor-recipient relationships by continen was the largest in the tropical groups (412), followed by sub-tropical (29), and temperat and boreal (16). The received amount was the largest in the tropical country grou (8881.21 million USD), followed by the sub-tropical group (90.25 million USD), and th temperate and boreal group (390.25 million USD). The fragmentation rate was the highes in the tropical group (22.57%), followed by the sub-tropical group (17.24%), and the tem perate and boreal group (12.50%).  Figure 1 shows the distribution of recipient REDD+ countries in which the x-axis i the number of donors and the y-axis is the fragmentation rate. Figure 2 displays the rela tionship between the fragmentation rate (y-axis) and the natural logarithm of the commi ted amount of finance (x-axis) among the recipient REDD+ countries. The statistics regard ing the amounts of commitments, the number of donors, and fragmentation rate of eac recipient country is summarized in Table 5. In both figures, it is assumed that the recipien countries whose fragmentation rate is higher than that of the total global REDD+ financ (21.88%) have a high degree of fragmentation.    From the concept of aid fragmentation in which too many donors are providing small amounts of aid, it can be indicated that, other conditions being equal, the relationship between the number of donors and the fragmentation rate is positive, and the relationship between the size of commitments and the fragmentation rate is negative. However, such relationships are not observed in the case of global REDD+ finance; no positive relationship between the number of donors and fragmentation rate is observed (Figure 1), and the correlation between the size of commitments and fragmentation rate is positive (Figure 2). These may indicate that the fragmentation of REDD+ finance has different characteristics from those of the typical aid fragmentation. Table 6 shows the fragmentation of bilateral and multilateral REDD+ finance from the perspective of donors. In total, 16 donor countries participated in bilateral REDD+ finance; on average, the number of recipient countries they supported was 17, the average amount of commitment was 387.36 million USD, and the fragmentation rate was 28.41%. Nine multilateral organizations participated in multilateral REDD+ finance; on average, 21 recipient countries were supported, the average amount of commitment was 389.07 million USD, and the fragmentation rate was 12.95%. Bilateral financing showed a higher fragmentation ratio than that of multilateral financing.

Classification of REDD+ Countries by REDD+ Progress
Of the 108 recipient REDD+ countries, 11 countries registered ERs and/or received RBPs (Group I). Thirty-six countries submitted FREL/FRLs and completed assessments (Group II). Fifteen countries registered their national focal points or were listed on the UNFCCC REDD+ website (Group III), and 46 countries did not submit (Group IV). The average number of donors and amount of allocated finance were highest in Group I and were second-highest in Group II. Group I and II showed a relatively high rate of fragmentation, at 21.71% and 24.74%, respectively. This may suggest that countries in Groups I and II have high emission reduction potential and a high level of national concerns on deforestation, so they attracted a large amount of finance from many donors, which made the financial aid in these groups relatively more fragmented. In Groups III and IV, the average number of donors, amount of allocated finance, and fragmentation rate were all lower than those in Groups I and II; this may suggest that countries in Group III and IV were relatively alienated from the REDD+ finance due to their lower competitiveness in emission reductions (See Table 7 and Figure 3).

Classification of REDD+ Countries by REDD+ Progress
Of the 108 recipient REDD+ countries, 11 countries registered ERs and/or received RBPs (Group I). Thirty-six countries submitted FREL/FRLs and completed assessments (Group II). Fifteen countries registered their national focal points or were listed on the UNFCCC REDD+ website (Group III), and 46 countries did not submit (Group IV). The average number of donors and amount of allocated finance were highest in Group I and were second-highest in Group II. Group I and II showed a relatively high rate of fragmentation, at 21.71% and 24.74%, respectively. This may suggest that countries in Groups I and II have high emission reduction potential and a high level of national concerns on deforestation, so they attracted a large amount of finance from many donors, which made the financial aid in these groups relatively more fragmented. In Groups III and IV, the average number of donors, amount of allocated finance, and fragmentation rate were all lower than those in Groups I and II; this may suggest that countries in Group III and IV were relatively alienated from the REDD+ finance due to their lower competitiveness in emission reductions (See Table 7 and Figure 3).

The Fragmentation of REDD+ Finance
The fragmentation rate was the highest in the tropical country group (22.57%), followed by the sub-tropical country group (17.24%), and the temperate and boreal group (12.50%) ( Table 4). A reason for this result could be that the deforestation rate is highest in tropical countries. As most tropical countries that have rainforests in these regions are developing countries in forest transition wherein deforestation occurs, the potential for emission reduction through REDD+ is very high; consequently, the larger the tropical forests, the higher the number of donors may be, leading to an increase in fragmentation rate.
The fragmentation of global REDD+ finance (21.88%) is lower compared to ODA aid fragmentation (40%), as reported by [32]. Considering that aid fragmentation of ODA occurs due to the small amount of finance that is delivered by many donors, this result can be attributed to the fact that the number of participants and the amounts of commitment in REDD+ finance are smaller than those of ODA. Besides, the existence of many recipient countries that received REDD+ finance from only a few donors may also have caused fragmentation to be underestimated. Twenty-seven recipient countries-one-fourth of the total-had only one donor-recipient relationship; although these recipient countries may not be significant from the perspective of the donors, the donors are significant for the recipient countries as they are the only channel for the REDD+ finance.
Among recipient countries, there was no clear relationship between the fragmentation rate and number of donors (Figure 1), and the relationship between the fragmentation rate and size of committed finance was positive (Figure 2). These are different from the explanations of aid fragmentation in which a higher number of donors and smaller size of commitments lead to more fragmentation. This difference may be explained by the existence of a few major donors who occupy a large portion of the total size of committed REDD+ finance. For example, Norway covers 22% of the total size of REDD+ commitments by itself; it also has committed more than 90% of funds to Guyana, making the other four donors' support relatively more fragmented.
The fragmentation of bilateral REDD+ finance (28.41%) is higher than that of multilateral REDD+ finance (12.95%) ( Table 6). Donor countries with bilateral relationships tend to consider not only the needs of the recipient countries but also their own potential profit when selecting recipient countries [31]; thus, finance may be herded into to a few favorable recipient countries that facilitate meeting the desired profits of developed countries, thereby increasing the fragmentation of REDD+ finance support. Although multilateral organizations support a greater number of recipient countries compared to bilateral REDD+ finance, the fragmentation of multilateral REDD+ finance seems to be lower as multilateral organizations support recipient countries with little or no support from developed countries.
It was observed that the majority of total REDD+ finance comes from a few "major" donors. For example, Norway and Japan were two of the major donors, and Norway has provided a large amount of finance to some recipient countries such as Indonesia and Brazil. This "inequality" among donors in terms of the provision of REDD+ finance makes the donor-recipient relationship with less finance involved appear relatively more fragmented, different from aid fragmentation in ODA, which occurs because small amounts of finance come from too many donors. In this regard, fragmentation in REDD+ finance does not necessarily have the same meaning as that of aid fragmentation defined by [32]. Considering that the amount of REDD+ finance has been insufficient to meet demands and that majority of recipient countries have had limited accessibility to REDD+ finance, the fragmentation of REDD+ finance may be understood as diversification of access to REDD+ finance for recipient countries. Thus, it may not be appropriate to consider fragmentation in REDD+ finance merely as an object that needs to be decreased.
Still, the total amount of REDD+ finance must be increased and adjusted for various donor-recipient relationships, in consideration of the REDD+ finance options in the Paris Agreement. To this end, an information-sharing system that enables transparent management of the financial commitments and the track of REDD+ countries' demands needs to be established first. The REDD+ partnership, in which 74 countries participated and interacted to share knowledge and experiences, and established VRD that collected information on REDD+ cooperation cases, including donor-recipient relationships, is one such effort. However, as the REDD+ partnership expired as of 2014, the existing VRD should now be updated. Also, while [1] states that the REDD+ implementation results for each relevant period must be reported, as well as the assessed FRLs/FREL, the SIS summary, a link to NSAP, and information on the NFMS, this reporting list does not include the information on REDD+ cooperation between countries, although it is important for the coordination of REDD+ finance. Therefore, such information needs to be collected and managed in the suggested information-sharing system.
Based on such a system, adjustments are required to achieve maximum effectiveness with a limited amount of REDD+ financial resources by ensuring that the small yet valuable financial support from smaller donors is not diluted by those of larger donors. Examples of such adjustments include pairing the REDD+ finance of smaller donors with those of larger donors, or functional differentiation such that larger donors support long-term and large-scale REDD+ programs, while smaller donors support short-term and small-scale REDD+ activities.

Adjustments under the Paris Climate Regime
Although we analyzed REDD+ finance committed under the Kyoto Protocol regime, since the Paris Agreement regime has started as of 2021, ways of adjusting fragmented REDD+ finance should consider aspects of the Paris Agreement such as NDC, cooperative approaches, and ITMOs.
Of the 108 countries that reported receiving REDD+ finance, only 12 countries have registered REDD+ results on the UNFCCC REDD+ web platform, which suggests that many countries are still in the REDD+ readiness phase or in an early phase of implementation, thus requiring additional supports to develop WRF. The focused allocation of REDD+ finance in countries in Groups I and II might have resulted in a high fragmentation rate in these groups. Considering the lack of REDD+ finance compared to overall demands and higher levels of REDD+ accomplishment of the REDD+ countries that received larger amounts of finance, the fragmentation rate can be interpreted as not only an indicator of the necessity of coordination but also an indicator of the accessibility of REDD+ finance to REDD+ countries. In other words, countries with a lower fragmentation rate have less accessibility to REDD+ finance, as shown in the analysis of Group III (those in their initial stage of national REDD+) and Group IV (those that did not fulfill WRF requirements). Considering the importance of emission reductions in the forest sector to meet climate targets, guaranteeing sufficient opportunities to participate in such efforts for countries in Groups III and IV through sufficient support of REDD+ finance should not be neglected.
Under the Paris Agreement regime, bilateral REDD+ cooperation through cooperative approaches to Article 6 is likely to be made not only to support WRF progress and provide RBPs but also to generate and transfer ITMOs. The REDD+ progress of recipient countries and objectives and motivations of donors and recipients need to be considered when adjusting the fragmented bilateral REDD+ finance. It would be desirable for REDD+ countries in phase I or II, which are developing WRF elements but have not registered any results yet, to receive support from developed countries or multilateral organizations that are willing to provide technical and financial supports to their recipients without any direct benefits (i.e., mitigation outcomes). REDD+ countries in phase III that have developed WRF elements and registered REDD+ results will attempt to receive RBPs and may seek additional support to enhance their existing REDD+ MRV system, as well as cooperative approaches with developed countries to scale up their national REDD+ activities. Thus, it would be desirable for developed countries that aim to encourage REDD+ countries to set more ambitious reduction targets and/or provide technical support to catalyze the improvement of existing MRV systems or to secure ITMOs from REDD+ countries to provide REDD+ finance to those in phase III. REDD+ countries grouped by REDD+ progress and possible REDD+ finance options are presented in Table 8.
Moreover, as multilateral organizations carry out REDD+ finance in accordance with their own objectives and criteria, they may conduct relatively less fragmented REDD+ finance compared to that of bilateral relationships under the Paris Agreement regime. To achieve climate justice and objectives, it is important for multilateral REDD+ finance under the Paris Agreement regime to play a safeguarding role for REDD+ countries that are relatively alienated from bilateral support.

Contributions and Limitations
This study contributes to the existing literature and the field at large by providing a detailed analysis of the fragmentation of REDD+ finance and specific recommendations to increase the effectiveness of REDD+ finance, which is crucial for REDD+ implementation in developing countries and consequently critical for achieving the goals of the Paris Agreement regime. However, there are limitations in this study owing to conducting analyses based on the data of REDD+ finances carried out under the Kyoto Protocol regime, due to the insufficient number of REDD+ finance investments conducted under the Paris Agreement regime. Further studies that track and monitor REDD+ cooperation cases under the Paris Agreement regime, especially those based on cooperative approaches, are required to contribute to the effective achievement of NDC and climate goals through REDD+. While we investigated the fragmentation of REDD+ finances only from the perspective of aid allocation, follow-up research could simultaneously consider the institutional arrangement of REDD+ finance under the Paris Agreement regime to analyze the fragmentation and make policy recommendations.

Conclusions
The importance of REDD+ for climate change mitigation and adaptation is expected to be increased in the Paris Agreement regime. Under the Kyoto Protocol Regime, various donors have committed REDD+ finance to support the implementation of REDD+ in developing countries. Voluntary and independent commitments without coordination have led to fragmentation and related inefficiencies in REDD+ finance. Therefore, we analyzed the fragmentation of REDD+ finance conducted under the Kyoto Protocol Regime and provided recommendations for coordination under the new Paris Agreement regime. The fragmentation of REDD+ finance among recipient countries was relatively reduced when compared to that of ODA, as the majority of the supported finance came from a few major donors. The fragmentation of bilateral REDD+ finance was higher than that of multilateral REDD+ finance, as developed country donors tend to concentrate supports to recipient countries that are favorable to the fulfillment of their own motivations.
Therefore, the fragmentation of REDD+ financing explains both the inefficiency of aid allocation, which can be improved through aligned supports, and the diversity of REDD+ finance, which shows accessibility to the support. Some REDD+ countries have made progress in national REDD+ and accomplished emission reductions; however, there are still many countries developing elements of WRF, and the other countries are in the very early stages of implementing national-level REDD+. Thus, REDD+ finance needs to be stratified in consideration of the progress of national REDD+.
Most developing countries are still in the REDD+ preparation phase and require more finance to increase their capacity. For these countries, aid for participating in REDD+ and developing and implementing elements of WRF can be delivered by donors, including developed countries and climate funds. For the REDD+ countries in phase 3, which have satisfied elements of WRF and registered the results, incentives can be delivered to encourage them to achieve more ambitious reductions. REDD+ finance for those purposes can be delivered by multilateral funds, such as the GCF, to reduce transaction costs and enable the coordination of REDD+ cooperation. Cooperative approaches can be another useful option for REDD+ countries in phase 3 and partner countries seeking ITMOs, enabling the REDD+ countries to secure resources to scale up national REDD+ support and the partner countries to achieve cost-effective mitigation.
For such forms of cooperation, an information-sharing and monitoring system that collects information on ongoing REDD+ cooperation, the commitment and expenditure of REDD+ finance, and the support needs of REDD+ countries at a global level should be established. Based on such a system, donors may provide aligned and predictable REDD+ supports and develop cooperation strategies that are functionally differentiated in consideration of their available amount of supports and WRF progress of the REDD+ countries. Multilateral organizations need to provide safeguarding functions for developing countries that are isolated from the bilateral REDD+ finance.

Data Availability Statement:
The data presented in this study are available on request from the correspondence author.