Introduction

Official development assistance (ODA) has traditionally been considered the main financial instrument to fight poverty. However, the effectiveness of this instrument has been widely scrutinised for decades, first from the perspective of the Global North (from innovative studies such as Dollar & Pritchett, 1998; Burnside & Dollar, 2004; Killick, 2004 or Easterly & Pfutze, 2008) to recent analyses such as Dreher et al. (2018) and later from the Global South (Almasifard, 2019; Van Dan and Binh, 2019; Lee et al., 2020; Mahembe & Odhiambo, 2017, 2020; Tang & Bundhoo, 2017). This criticism has opened the door to an extensive variance of concepts, such as Global Policy Finance (Severino & Ray, 2009), Beyond Aid (Carbone, 2012; Janus et al., 2015) or Beyond ODA (Xu & Carey, 2015), that have been calling for broadening the frame of reference when considering the instruments necessary to promote development processes.

The decline in ODA flows experienced by most traditional donor countries after the financial crisis of 2007–2008 (Hynes & Scott, 2013; Phillips, 2013) only reinforced this trend, as some countries called for the consideration of new mechanisms in the aid system (Ocampo, 2016). As a consequence, a wide range of financial instruments, such as philanthropic donations or remittance flows, has been subsequently proposed to be considered promoters of development processes (Adam et al., 2015).

Since the adoption of the 2030 Agenda, there has been a living debate on whether these instruments should be considered to foresee what the new global aid system could be like to achieve the sustainable development goals (SDGs) (King, 2016). Today, authors such as (Calleja & Prizzon, 2019; Pekmezovic, 2019) have already observed a paradigm shift from an aid-centric approach to financing for development (FfD) to a more holistic approach linking public and private resources.

The centrality of the SDGs as the guiding principles of development processes is today a reality, and the commitment of all countries to the 2030 Agenda seems unstoppable, even despite the recent COVID-19 crisis (Santos-Carrillo et al., 2020). However, there is still no agreement on the relative importance of each financial instrument when evaluating each country’s commitment to the 2030 Agenda (Sianes, 2017). There is no agreement even on which financial instruments should be considered as a part of this framework (Engberg-Pedersen, 2020).

The most recent proposal in this sense has been the Total Official Support for Sustainable Development (TOSSD) (Rogerson & Kharas, 2016). Designed by the Organisation for Economic Co-operation and Development (OECD) in the context of the Addis Ababa Action Agenda, this new framework calls for monitoring official resources and private finance to account for achievements in sustainable development. Still pending the TOSSD’s positioning as the go-to metric, the first comprehensive data set clearly shows how donor countries support diversified FfD approaches (Tomlinson, 2021). There is a wide range of national strategies when promoting development assistance, relying more on one or another of these instruments.

Although in recent years the countries involved in the aid system have dedicated great efforts to coordinating their actions, which has been traditionally deemed a lever of good results for developing countries (Fukuda-Parr, 2012), the FfD debate cannot be treated as collateral (Garcia-Arias, 2015; Griffiths, 2017; Kharas & Rogerson, 2016). If it is not appropriately addressed, it could lead to a situation of indeterminacy similar to that experienced during the 1990s, which could hinder development processes started during the last decade and reduce the effectiveness of the 2030 Agenda.

In fact, most recent statements in the arena of international development focus mainly on the issue of financing, especially since the outburst of the COVID-19 pandemic. International organisations have been monitoring the evolution of diverse funding sources, noticing a steady decrease in most of them despite remarkable public and private efforts (Sachs et al., 2022). Relevant leaders and academics (Goodell, 2020; Stiglitz, 2020) also demand a renewed effort to support the implementation of the 2030 Agenda, calling for a shared and committed FfD agreement.

In this context, the purpose of this paper is to identify the various FfD strategies that donor countries are currently putting into practice in the global fight against poverty, under the assumption that it is possible to find groups of countries characterised by internal homogeneity (countries in a group or cluster follow similar strategies) and external heterogeneity (each cluster follows a different strategy). The analysis will thus address the following research questions: are different groups of countries sharing a similar strategic approach when promoting development worldwide? Could some of these approaches conflict when trying to reach a new international consensus on how to finance development?

This analysis will contribute to demonstrating how donor countries pursue different strategies, which can hinder the achievement of the necessary consensus to ensure effective financing of aid and ultimately, the attainment of the SDGs of the 2030 Agenda. The contribution to the literature in the field is twofold: on the one hand, to demonstrate through quantitative analysis the diversity of development financing strategies followed by countries; on the other hand, to reflect, based on the literature on the strategic analysis of industrial sectors and the game theory, how this diversity is an obstacle for countries to reach the necessary consensus to finance the 2030 Agenda.

To achieve this goal, the paper is divided into five further sections following this introduction. In the section “Theoretical framework”, we will introduce the theoretical framework: there is an emergent new global aid system comprised of elements that can be classified according to their governance (public versus private funds) and approach to development (global versus sectoral) into four categories that we will use to identify the different FfD strategies. Section “Methodology” formulates the hypotheses of the study: there are not only substantial differences between the FfD strategies carried out by each country, but it is also possible to find similar behavioural patterns in different groups of countries. To determine the clusters of countries that base their FfD strategies on the same instruments, we will use a multivariate analysis technique: hierarchical cluster analysis. Section “Results” presents the results: we identify four different clusters of countries according to the financial instruments they rely on to support international development processes. To do so, we introduce a graphic tool: the Aid Strategy Diamond. We discuss these findings in the section “Discussion of the results: the FfD Strategy Diamond”, and their implications in the section “Implications”. Finally, the article highlights the main consequences that can be drawn from these results in the section “Conclusions, limitations and future research” on conclusions.

Theoretical framework

The underlying theory behind this research is rooted in a basic game theory assertion: it becomes more difficult to cooperate as the number and the heterogeneity of players increase. The rational approach to international relations adopted this assertion: co-operation (in a very broad sense, for example, coordinating on the most effective financing instruments) among countries is more difficult when the preference heterogeneity is high (Koremenos, 2016). In this case, and according to this author, there is a “Coordination but not distribution” problem because agreeing or not on a global FfD strategy requires coordination, but there are no gains upon which countries have to bargain. Coordination should not be understood as an “all-or-nothing problem”, because “there is no need to meet exactly, provided that policies fall within a range of compatibility.” (Snidal, 1985).

The issue here is that the mere existence of groups of countries following different FfD strategies will make it more difficult to reach a global agreement on how to finance development, for two reasons. First, this strategic heterogeneity is the evidence of different (and perhaps irreconcilable) ways of understanding co-operation, and particularly, the role that the state should play in it. A second reason can be linked, with all necessary caveats, to the literature on business strategy. Groups of countries with similar FfD strategies can be equated to the concept of a strategic group, a set of companies that follow similar strategies. The more diversity there is in the strategies pursued by these groups, the more difficult it is to cooperate, because “these differences will complicate the process of firms understanding each other’s intentions and reacting to them” (Porter, 1980).

Financing for development in a beyond-ODA aid system

Numerous scholarly contributions, ranging from early works (Severino & Ray, 2010; Severino & Ray, 2009) to more recent studies (Leach et al., 2021; Sianes 2021), reflect a general acknowledgement within the academic community regarding the substantial transformations observed in the development aid system over the past decade. The world has suffered the consequences of the profound 2007–2008 financial crisis, which sprung from and mostly affected developed countries for the first time in recent history (Woods, 2010). The recent COVID-19 pandemic has exacerbated the negative impacts of the financial crisis, especially for less developed countries (Oldekop et al., 2020). Meanwhile, some developing regions have continued their emergence, as they have suffered the crisis with lesser intensity. In this changing context, the new Development Agenda and the FfD debate can be interpreted (Leach et al., 2021).

In recent years, the international community has made remarkable efforts to address some flaws of the traditional aid system, as manifested in the adoption of the 2030 Agenda. However, countries involved in this process have not yet agreed on many other topics, such as the most appropriate instrument to finance such an Agenda. The main debates about this topic have gone in two directions: how to extend the public funds mobilised from co-operation policies to a broader spectrum of public policies, and how to best mobilise private resources for development (United Nations, 2021).

Although some countries (the United Kingdom and Germany being the most remarkable) have increased their ODA commitments, since 2009 we have witnessed signs of exhaustion in ODA flows from most traditional donor countries (Nilima Gulrajani & Swiss, 2018). At the same time, other financial instruments emerged, drawing a more comprehensive financial system for development (Zimmermann and Smith, 2011; Nunnenkamp and Thiele, 2013). These instruments going “beyond” traditional development aid formulas were already shaping the FfD scenario during the negotiation of the 2030 Agenda (Michael, 2016; Pekmezovic, 2019).

To respond to the increasing analytical needs in this area, the OECD Development Assistance Committee (DAC), a forum with 31 of the largest aid providers, has been working to improve the quality of its statistics on financial aid instruments to developing countries beyond ODA. These instruments include flows for development mobilised from the private sector, private philanthropy, Foreign Direct Investment or remittances. Apart from those, there is a complementary approach called the policy coherence for development (PCD), which does not channel financial resources but seeks to enhance the complementarities of all public policies in a donor country. Our thesis is that, in general terms, these instruments can be classified according to two criteria.

Main financial instruments in a beyond-ODA scenario

To classify the financial instruments for development, the first criterion refers to the governance of the instrument in terms of whether it is public or private. Both types of instruments are affected by public policies and executed within the borders of each country, but while the former are under the direct control of donor governments, the latter are finally executed by private actors. The second criterion is their approach to development, to distinguish whether they are focused on certain development co-operation sectors (sectoral) or aim to support overall development processes that are not sector-specialised. This classification gives rise to four categories of FfD instruments, as the Cartesian product of both criteria (Fig. 1).

Fig. 1: Categorisation of FfD instruments.
figure 1

This figure depicts a classification of the different FfD instruments according to their governance (axis Y) and their approach to development (axis X).

Full size image

For the purpose of this study, we have chosen one instrument per category as a proxy of its behaviour, namely, classic ODA flows, the PCD approach, remittance flows and private philanthropic donations, which relate to each criterion, as shown in Table 1.

Table 1 Finance for development instruments by criteria.
Full size table

We assume that these four instruments deserve more consideration than others when designing the new global FfD Agenda. As we will show, they have been deemed to deeply influence the aid system. Two additional reasons support the focus on these four instruments in conducting our empirical study: first, we want to analyse the behaviour and strategies for reducing poverty promoted by donor countries. This empirical analysis will not consider aid instruments led by multilateral institutions, recipient countries, or companies and enterprises (normally transnational). Second, we want to evaluate countries’ public commitment to development, so we will only take into consideration instruments that can be directly executed or indirectly promoted by national public policies.

Thus, the four instruments and the indicators used to measure their performance in this paper are the following:

Official development assistance (ODA) flows: This is the term coined by the OECD-DAC to measure aid flows, which satisfy three conditions: they are undertaken by the official sector; the main objective must be the promotion of economic development and welfare, and they must be approved under concessional financial terms (OECD, 2021c). We classify them as a public instrument because donor countries’ governments make relevant decisions about their quantity, allocation and purpose. We also consider them sector-oriented, as these decisions are specifically taken within the development co-operation system. Today, the measurement of ODA is accurately defined by the OECD-DAC. This institution has been collecting data since the 1960s, and today, it publishes data and analyses each year for every country in the world. ODA volumes may be measured absolutely, by the amount transferred, or relatively, as a proportion of the donor-country’s size. For the statistical analysis, this research will use the percentage represented by ODA flows in each donor-country’s total global national income (GNI) for 2019.

Policy coherence for development (PCD): This approach states that it is necessary to enhance complementarities between the various policies of a donor country to achieve greater sensitivity and guidance for the development goals of all public government policies (Alonso et al., 2010; McLean Hilker, 2004; Sianes, 2017). Therefore, we classify it as global-oriented, as the PCD approach and the decisions made within its context do not focus strictly on specific development co-operation processes; we consider it a public instrument, as all decisions are promoted and executed by donor countries’ governments. Remarkably, recent studies underline how PCD is essential for designing and implementing the 2030 Agenda (Brand et al., 2021; Koff et al., 2020). At the present moment, the most accepted and accurate way of measuring PCD is the Commitment to Development Index (CDI) developed by the Centre for Global Development (Birdsall & Roodman, 2003). The CDI is a complex index that assesses the commitment of rich countries in seven different policy areas: aid, finance, technology, environment, trade, security and migration. In each component, scores are scaled so that 5 is the average because there is no such thing as an international agreement regarding the accurate level of coherence. This paper will use the 2018 edition but remove the aid component to avoid doubling the ODA effect.

Philanthropic flows (PHI): technically, a philanthropic donation is the transfer of money from “individual and institutional donors to support charitable causes across national borders” (Indiana University Lilly Family School of Philanthropy, 2020). In the field of development aid, philanthropic flows are understood today as investing resources to create sustained solutions, focused not on the recipient organisation itself but on a solution to end the development need (Peterson, 2015). Diverse research studies have emphasised the role of these flows in the pandemic context (Godwell et al., 2020; D. Walker, 2020). Although tax regulations (Neumayr & Handy, 2019) and other public decisions (Wiepking, 2021; Wiepking et al., 2021) affect philanthropic flows (this is why we include them in our analysis), the final governance of the flows is private (Moran & Stone, 2016), and we can consider them sector-oriented on development needs. These cash flows are not easily identifiable since there could be undeclared private donations that are hard to identify without proper accountability tools. However, different institutions examine the involvement of countries in philanthropic flows. In this case, we will develop the analysis in accordance with the Global Philanthropy Tracker data by Indiana University Lilly Family School of Philanthropy, which captures the global cross-border individual and institutional philanthropic contributions. Our data reflect the philanthropic outflows as a share of GNI of each donor country.

Remittances (REM): Remittances are the personal flows of money from migrants with destinations in rich countries to friends and relatives in their poorer countries of origin. Ratha (2007) stated that they impact development, as they improve the financial access of migrants, their beneficiaries, and financial intermediaries in the countries of origin. Recent research studies verify how remittances are growth-enhancing (Cazachevici et al., 2020) and particularly important for financial development in developing countries because they foster long-run growth and reduce poverty (Azizi, 2020). However, we can consider remittances more global-oriented, as they are not always directed to specific development uses. Additionally, public policies making remittance services cheaper and more convenient could indirectly leverage these flows, but at the end of the day, they also have private governance. Today, different institutions rate the performance of countries in remittance flows. In this case, we will perform the analysis using World Bank data on bilateral remittance flows. To measure remittances to developing countries properly, we have filtered data, so we only consider flows from DAC countries to non-DAC countries. The most recent data with this information are from 2018. The resulting amount has been divided by the total population of each donor country for 2018 (United Nations, 2018). The reason to do so assumes that a way in which developed countries can support developing countries is by facilitating the immigration process, including the sending of remittances to their families. The effort that a recipient-developed country exerts in accepting immigration and allowing remittances depends, to a great extent, on its population. The same amount of remittances from a larger or a smaller country in terms of population would mean different levels of tolerance to immigration and thus emission of remittances.

Table 2 summarises the FfD instruments considered and the indicators used for the analysis. The year selection is based on the most recently available data at the time of retrieval.

Table 2 FfD instruments and chosen indicators.
Full size table

Relying on empirical data opens the possibility to test some relevant hypotheses, filling some of the above-identified gaps in the literature. However, the information presented above speaks of the limitations of available data, not only in terms of geographical and historical coverage but also on representativeness, as they can only be used as proxies of the different financing strategies. Such limitations impact the scope of the research by constraining potential analyses. For example, no instrument but ODA allows for analyses on recipient countries, and most of these instruments have been developed only recently, so longitudinal analyses are also limited.

Nevertheless, these limitations do not impede the essence of our research, as our primary focus lies in donor strategies and the present context, aiming to ascertain whether these disparities hold sufficient significance to account for the challenges encountered in achieving a worldwide consensus on financing the 2030 Agenda.

Methodology

Justification and hypotheses

The Introduction has stated how the development aid system has changed in recent years. Former agreements on goals, actors and instruments have been questioned, and the ratification of the 2030 Agenda has only pushed the need for new international consensus on other topics, such as on FfD.

The importance of our analysis relies on the necessity of an agreement to set up a shared FfD agenda among donor countries, as today’s agenda narrowly focuses on ODA flows. Two facts demonstrate the relevance of this debate. Firstly, the Sustainable Development Report 2023 states that “The international financial architecture is failing to channel global savings to SDG investments at the needed pace and scale.” Secondly, the convening of The Summit for a New Global Financial Pact in June 2023, calling for increased agreement among countries to finance aid, further highlights this issue.

The most significant contribution of our article is to demonstrate that donor countries pursue different strategies, which can hinder the achievement of the necessary consensus to ensure effective financing of aid and ultimately, the attainment of the SDGs of the 2030 Agenda. To this purpose, we test two hypotheses:

Hypothesis 1: there are different groups of countries that share similar FfD strategic patterns.

Hypothesis 2: the FfD strategic patterns mentioned above diverge in the weight of the financial instruments countries use to fund development.

Highlighting the differences among groups of countries will identify the reasons behind the difficulties in reaching an agreement on a new shared FfD agenda, so identifying these clusters of countries may help find potential partners when negotiating the incoming FfD agenda.

Statistical procedure and data

To test these hypotheses, we perform a cluster analysis. Cluster analysis is a statistical method of partitioning an observed sample population into relatively homogeneous classes. In this case, we want to classify a sample of countries into a number of groups or clusters so that the degree of statistical association is high among members of the same group and low between members of different groups (Berlage & Terweduwe, 1988).

Cluster analysis has been successfully performed with small samples of countries in many research projects (Franzoni, 2014; Jurun & Pivac, 2010; Paap et al., 2005; Serrano et al., 2020; Sianes & Ortega Carpio, 2014; Vázquez & Montellano, 2012).

Authors recommend using hierarchical methods to perform the classification (Aaker & Day, 1989), where the researcher must decide on the number of clusters (Díaz de Rada Igúzquiza, 1988; Martínez Ramos, 1984). To determine the number of clusters on a sound basis, we applied the R package NbClust (Charrad et al., 2014) using 30 different methods. Hierarchical cluster analysis is a suitable method when the number of observations is less than 30, as is the case here (McNeish & Harring, 2017). Other clustering techniques, such as Latent Profile Analysis, require a significantly larger number of observations to provide robust results (Spurk et al., 2020).

Aldenderfer & Blashfield (1984) established four basic stages in carrying out a cluster analysis:

  1. 1.

    Identification of the subjects of analysis and the variables to be included.

  2. 2.

    Choosing a cluster classification method to create similar groups.

  3. 3.

    Calculation of distances or similarities between cases.

  4. 4.

    Validation of the test results.

Regarding the identification of the subjects of analysis and the variables (instruments in this case), this paper will analyse the FfD strategies of twenty-six DAC-donor countries, those for which there is information available for the four indicators: Australia, Austria, Belgium, Canada, Czech Republic, Denmark, Finland, France, Germany, Greece, Hungary, Ireland, Italy, Japan, Korea, Luxembourg, New Zealand, Norway, Portugal, Slovak Republic, Spain, Sweden, Switzerland, The Netherlands, United Kingdom and United States.

Table 3 summarises the collected data (which will be later normalised to homogenise the results and facilitate the graphic representation provided in the section “Results”).

Table 3 Data collected for each country.
Full size table

We used Ward as the clustering method (a hierarchical aggregation procedure) with the squared Euclidean distance, which is appropriate for our number of cases (Jain & Dubes, 1988). Once the clusters are formed, we can graphically represent them in a dendrogram, which displays the clustering process of observations. A dendrogram can be read from left to right (showing how the algorithm has grouped the observations) or from right to left (providing interesting insights about the order in which differences between great groups of countries emerge).

The last stage of the cluster analysis is the validation of the results. Apart from the classification of countries by itself, it is also interesting to test the hypothesis that the different behaviours found in each group of countries are statistically significant, and to analyse which variables are dominant in the formation of clusters. To do so, we respectively conduct a one-way ANOVA test and a Bonferroni test. We also calculate the Silhouette Coefficient (Rousseeuw, 1987) and Dunn’s Index (Dunn, 1974) to check the internal cohesion and separation of the clusters. Higher values indicate better clustering quality. Finally, we performed the k-means clustering process to check if the allocation of countries to clusters remains the same. To calculate the Silhouette Index and the k-means clustering, we applied the R package cluster, and to calculate Dunn’s Index, we used the R package clValid.

Results

The application of the R package NbClust determines an optimal number of four clusters, as shown in Fig. 2.

Fig. 2: Optimal number of clusters.
figure 2

This figure represents the resulting frequency for each number of clusters according to the different calculations methods. A higher bar represents more methods suggesting a given number of clusters.

Full size image

The resulting dendrogram is presented in Fig. 3. Read from left to right, we can see, for example, that in the first stage, individual countries merge to form 10 clusters, such as the group composed of Italy, Spain and Austria. Read from right to left, the initial division among countries forms two major clusters. The first cluster (on the top) consists of countries that, in general, allocate little effort to aid (whether public or private), while the second group (on the bottom) comprises countries that make a greater effort towards that purpose. Following this, each of these two groups is further divided into two, resulting in the four clusters previously determined as optimal.

Fig. 3: Dendrogram.
figure 3

This figure represents the resulting dendrogram using Ward’s method with squared Euclidean distance. From left to right, the lines represent when each cluster is conformed.

Full size image

Table 4 displays the composition of the four clusters and the average value for each indicator (centroids).

Table 4 Cluster membership and centroids.
Full size table

The results of the ANOVA (Table 5) and the Bonferroni tests (Table 6), which in this case analyses how each instrument contributes to generating differences between clusters, reveal that both hypotheses are proven. First, there are significant differences between clusters. Therefore, the behaviour of their belonging countries is significantly different (ANOVA). Second, the four instruments play a significant role in forming clusters (Bonferroni test). For example, PHI is different in clusters 1 and 3, 1 and 4, 2 and 3, 2 and 4, and 3 and 4. Both contrasts were performed at a significance level of α = 0.05.

Table 5 Results of ANOVA test.
Full size table
Table 6 Results of Bonferroni test.
Full size table

The Silhouette Coefficient is 0.39 (Cluster 1, 0.56; Cluster 2, 0.47; Cluster 3, 0.27; Cluster 4, 0.24) and the Dunn’s Index is 0.3125. Both values indicate that the clusters have an acceptable internal cohesion and separation from the rest.

The k-means clustering procedure classified the countries in exactly the same clusters, so we can conclude that the clustering we have performed is robust.

Discussion of the results: the FfD Strategy Diamond

According to the ANOVA and Bonferroni test results, we should consider the four instruments to explain the differences between clusters. To help understand these results, we will accompany each cluster definition with a graphic.

The graphic is designed as a diamond, using the standardised values of the indicators: the right axis reflects the PHI indicator; the lower one, REM; the one on the left shows the score for PCD; finally, the upper axis reflects ODA flows.

Figure 4 represents the mean scores obtained by each cluster in the four instruments. We discuss the main characteristics of each cluster below.

Fig. 4: The FfD Strategy Diamond for the four clusters.
figure 4

All remaining figures represent the FfD Strategy Diamond. Each axis represents one of the four considered variables, and each figure depicts average results in each variable for each cluster. The longer the value in each axis, the more present is such FfD instrument in each cluster. In this case, average values for the whole sample is represented.

Full size image

Cluster 1: “The Underachievers”

This group comprises seven European Union countries: Austria, Belgium, Finland, France, Italy, Portugal and Spain. As shown in Fig. 5, these countries share similar figures in every type of instrument, being PCD their main strategy with a high average value of the indicator, 5.23. However, they show relatively low scores in the other FfD instruments: ODA (0.31%), REM (156.57 US$/person) and especially PHI (0.03%). Given their location in the European Union and size, these countries could play a more relevant role in the international arena regarding FfD, so we call them “The Underachievers”.

Fig. 5: The FfD Strategy Diamond for cluster 1.
figure 5

This figure represents average values for countries comprising cluster 1. To check the average result of cluster 1, please check fig. 4.

Full size image

The main risk for all these countries is that they can lose (if they ever had) relevance in the development aid debate if they do not develop a solid position on FfD. At the moment, they are not very relevant in the debate on climate change and sustainable development (Li, 2009; de Noronha and Vaz, 2019; Szopik-Depczyńska et al., 2017), so they need to be aware of that and take advantage of the discussion between clusters 3 and 4 concerning public or private flows to put their interests on the development aid agenda.

These European countries, especially those from the South (Spain, Italy and Portugal), present some limitations today since they face societies that are quite worried about the financial crisis and the impact of the pandemic, so public support regarding ODA will be at risk, as is already happening in some countries (e.g., Spain ODA as a share of GNI fell from 0.43% in 2010 to 0.12% in 2015, stabilising at approximately 0.20% from 2017 to 2020 (OECD, 2021a). On the other hand, their citizens are less accustomed to large private donations (Wiepking, 2021; Wiepking et al., 2021). This combination of factors explains why “The Underachievers” struggle for a development agenda based on PCD.

In this regard, European countries can display some common policies (environmental and commercial, mainly) that are already considering the interests of the poor. The PCD approach could find support and renewed enthusiasm in their societies, as it has been one of the main demands of academics and Northern civil society organisations in recent years.

Cluster 2: “The Outsiders”

This group comprises two Asian donor countries, Japan and South Korea, and some European countries: Greece, Czech Republic, Hungary and Slovak Republic. As Fig. 6 shows, there are very slight differences among countries, as they present the lowest levels in every single FfD instrument, either public or private: ODA, 0,18%; PCD, 4,60; PHI, 0,02% and REM, 72,23. For their lack of prominence in the international co-operation system, we call them “The Outsiders”.

Fig. 6: The FfD Strategy Diamond for cluster 2.
figure 6

This figure represents average values for countries comprising cluster 2. To check the average result of cluster 1, please check fig. 4.

Full size image

Greece, Japan and South Korea are almost invisible in the graphic. This situation could make sense for South Korea, which joined the donor club in 2009 (Kim, 2011). However, what is happening with the rest needs to be explained.

Japan has always lived a dynamic relationship with the development aid system. For example, ODA flows have traditionally been deeply influenced by their own economic cycles (Cooray et al., 2005; Yasutomo, 1989), and the initial interest shown in the Commitment to Development Index (Sawada et al., 2004) soon evaporated. It reflects how Japan’s aid policy is mostly a continuation of its domestic postwar economic recovery strategy, that is, a more evident concern for its domestic prosperity and security (Yamamoto, 2017).

Regarding the European countries within this cluster, the Greek performance can be explained according to their most recent history: an enormous public debt that has pushed them to abandon their ODA commitments, and no private finance has still filled the gap. On the other hand, the Czech and Slovak Republics show very similar behaviour, with scarce attention to development issues in their public policies apart from a slight commitment to PCD, mainly due to the influence of European policies.

As a consequence, we cannot easily anticipate the position of “The Outsiders” in the incoming FfD debate. Their relative irrelevance in current FfD instruments might push them to adopt an old-style approach to development co-operation closer to Foreign Direct Investment (Adams & Opoku, 2017; Kragelund, 2011). In this regard, they could be tempted to join forces with BRICS countries (O’Neil, 2001), deepening the shift in global economic power from Western countries towards such emergent economies: Brazil, Russia, India, China and South Africa.

Cluster 3: “The Traditional Winners”

It is logical what to expect from the seven countries that belong to cluster 3: Denmark, Germany, Luxembourg, Norway, Sweden, the Netherlands and the United Kingdom. As Fig. 7 shows, they share a development strategy based on the highest levels of ODA (0.82%). Actually, in this group, we can find the only countries that have traditionally accomplished the promised 0.7% ODA/GDP. For this reason, they are considered “The Traditional Winners”.

Fig. 7: The FfD Strategy Diamond for cluster 3.
figure 7

This figure represents average values for countries comprising cluster 3. To check the average result of cluster 1, please check fig. 4.

Full size image

All but Germany and the United Kingdom are small European countries that have suffered financial crisis effects at lower levels (Rose & Spiegel, 2012), and they are well-developed welfare states.

An aid strategy based on this configuration of instruments would generally present the opposite characteristics to those of cluster 4, “The Champions of Philanthropic Flows”. The main advantage could be the higher predictability and easier governance of ODA flows.

However, this approach to FfD currently faces three main criticisms. First, the drawbacks of the administrative costs it generates (Birdsall & Kharas, 2010); second, the ODA approach is suffering a period of crisis since it has not proven to be effective enough in the fight against poverty and inequality in the last 50 years (Bird & Choi, 2020), and third, currently, public deficits are the main concern of most donor countries, so citizens will be especially reluctant to keep backing this approach while their welfare state is being eroded (Gulrajani & Faure, 2019).

In the context of the negotiations to determine the new global agenda for development aid, their homogeneity can be a strategic advantage, as they will probably argue for the continuity of ODA as the main instrument in the fight against poverty. Another factor to consider is that they are powerful within the European Union, so they will probably try to persuade “The Underachievers” to join efforts with them, as they have already done in other international issues (Gómez, 2018).

Cluster 4: “The Champions of Philanthropic Flows”

Six countries comprise cluster 4: Australia, Canada, Ireland, New Zealand, Switzerland, and the United States. Most of them are Anglo-Saxon countries, whose tradition relies on individual responsibility. This group shows (Fig. 8) the highest scores on private donations, both from local people and institutions (PHI, 0,13%) and immigrants (REM, 364.03), whereas ODA and PCD score are relatively low (0.28% and 4.89, respectively). The high correlation between the two instruments of private giving makes sense in societies that value peer-to-peer solidarity, so we call them “The Champions of Philanthropic Flows”.

Fig. 8: The FfD Strategy Diamond for cluster 4.
figure 8

This figure represents average values for countries comprising cluster 4. To check the average result of cluster 1, please check fig. 4.

Full size image

The main advantages of this approach to FfD are the following: first, the reduction of transactional costs, which is one of the main concerns of citizens when making donations to aid organisations and institutions; second, private aid is usually addressed to citizens and families in developing countries (remittances) and to civil society organisations in the South, which promote social inclusion and entrepreneurship, and third, funds are caught with agility, which is essential, for example, when natural disasters happen.

On the other hand, the main drawbacks are that cash flows are not predictable, which has been one of the main demands from recipient countries since the Paris Declaration on aid effectiveness; and the governance of the aid system would be even harder since high amounts of funding could be concentrated in some regions while others keep unattended.

In the incoming FfD debate, these countries are likely to struggle for an aid system strongly based on private giving. This movement will be more significant after the COVID-19 crisis. In the context of fighting against public deficits, reducing the burden of higher levels of ODA would be a solid strategy, perhaps followed by many other donor-country governments all over the world.

Implications

Today, there is a profound FfD debate that questions whether to use instruments other than ODA or not. The lower support received by newer FfD instruments hinders their technical development, affecting their international acceptance and thus their dissemination levels. Consequently, these other instruments are not yet consolidated nor adequately systematised. However, the results show how three of the four clusters do prefer them to ODA, which has implications for all actors in the international system.

At the global level, international organisations leading the development system, such as the OECD-DAC, must be aware of this new reality. Even though the recently developed TOSSD represents a step forward in recognising such diversity, consensus is still needed. Other multilateral organisations, in turn, as receivers and channelers of aid, can contribute to dampening these divergences by integrating alternative sources of funding and by tailoring their accountability, developing more robust data management systems.

At the national level, the analysis has implications for both donor and recipient countries. The former might use this analysis to find potential allies in strategic debates, or at least to identify possible strategies to emulate. But they also can identify their current belonging to a certain strategic group as unsatisfactory, thus acting accordingly. The latter must be aware of the diverging financial approaches of their donor allies, especially if they receive support from donor countries belonging to different clusters. The mere coexistence of different financial logics might have an impact on recipient countries’ development processes.

Conclusions, limitations and future research

This paper shows how today coexist different approaches in terms of FfD among countries, rendering it possible to clearly identify groups or clusters of countries, each of which shows a relatively common FfD strategy that is significantly different from those displayed by the rest. As shown, the four clusters identified have different interests and approaches to the most suited financing strategy to achieve poverty reduction. One of them, “The Champions of Philanthropic Flows”, comprising Anglo-Saxon countries, focuses on private giving. On the other hand, “The Traditional Winners”, a cluster mainly composed of European Nordic countries, clearly prioritises ODA as the primary means of financing development aid. Finally, Western European and Australasian countries are not so specialised: some of them, “The Underachievers”, protected by European Union institutions, narrowly focus on PCD, while “The Outsiders”, such as Japan and South Korea, play a marginal role in the current system.

Given the inherent uncertainty in countries’ preferences, predicting the stance each cluster would take in a potential debate on FfD is difficult. However, it is unlikely that there would be extreme shifts in positions. For example, it is difficult to imagine that the United States would cease prioritising private instruments and give greater relative weight to ODA. Conversely, due to their historical tradition, it is also unlikely that Nordic countries or the United Kingdom would relinquish their role as major ODA contributors.

Such divergence only reflects the most recent tendencies on global governance, according to which some authors claim that we are in a transitional period of both polarisation and a certain lack of leadership. In this context, the United States, the European Union and Japan are paralysed by their own internal problems, and Multilateral Financial Institutions do not have sufficient consistency to take the lead. According to our analysis, this state of paralysis is also evident in terms of financial strategies, as it demonstrates how various approaches and development aid strategies coexist today, with none being superior.

Despite the clarity of the results, the empiric nature of this study has some limitations. First, regarding data availability, validated information on the different instruments has restrictions regarding periodicity and content. Not all donor instruments can be related to recipient countries nor segmented by sector or activity. However, relying on quantitative data opens the possibility to test relevant hypotheses empirically. Second, the data retrieved only allows for a cross-sectional analysis. In contrast, a longitudinal analysis would allow us to study the dynamics over several years to see whether the variations overlap with the socioeconomic concerns of the donor countries. Nevertheless, we have performed a qualitative dynamic interpretation of the clusters in the discussion section to cover this gap partially. Finally, the elicitation of diverse funding strategies does not provide information on their effectiveness.

These main limitations, in turn, imply future avenues of research. First, incorporating more sophisticated data as it becomes available would open the door to a more thorough donor-recipient analysis. Recent developments suggest that the TOSSD methodology can be helpful in this regard. Second, adding more updated data cohorts will provide a dynamic lens on how the different clusters are evolving, if so, and which countries are de facto leading each of the identified approaches. Both improvements could provide valuable insights into the effectiveness of each financial strategy.

Lastly, another possible future line of research would be to compare the strategies of these countries with those that are not part of the DAC.

If the international community wants to “leave no one behind”, it is time to go deeper into the necessary debate about how to finance the 2030 Agenda. This paper may contribute to this debate by clarifying one of the possible sources of the lack of agreement, namely, the existence of diverging strategies followed by donor countries.