The AIDSOCPRO project – Aiding Social Protection – focuses on the political economy of externally financing social policy in developing countries. External financing in this sense refers to official development assistance (aid for short) and other official flows, such as non-concessional official lending from international financial institutions (IFIs). The project examines such financing in support of social protection programmes in developing countries, mainly focusing on cash transfer (CT) programmes and related targeting systems.
The problems that we have been exploring are two-fold.
In bringing together these two strands of research, we hypothesised at the beginning of our project that the increasing emphasis of social protection by donors would tend to exacerbate the politicised controversies and power struggles already apparent within donor-recipient relations, as well as within the recipient countries themselves, from macroeconomic conditionalities to negotiations regarding domestic spending commitments and even the design of social protection programmes. This contrasts to most of the scholarship on the politics or political economy of social protection, which takes external financing as non-problematic and mostly focuses on domestic matrices of elites in order to understand variations in the uptake and implementation of cash transfers within and across countries.
'The implications of these questions require a serious rethink of many of the accepted premises in the political economy of aid'
On the latter point, our approach also differs from this dominant focus on variation. From a social policy perspective, we contend that there is actually a strong uniformity in the broad design of CTs across developing countries, despite differences in patterns of precise uptake and implementation. Hence, we have been interested in asking why this uniformity has occurred and, if we accept the role of external influence, how has this influence occurred such that relatively uniform outcomes have been achieved across a wide variety of settings.
The implications of these questions require a serious rethink of many of the accepted premises in the political economy of aid and related literatures, in particular highlighting the often-contradictory tensions that recipient governments must deal with in order to preserve policy space in the face of various international development agendas and aid modalities. They also help us to re-examine our understanding about causality and motivations regarding the adoption and expansion of CT programmes, and the politics around recipient government ‘ownership’ of such programmes.
Emma, Ben, Ana, and Charmaine in the panel that they convened with Andrew. at International Initiative for the Promotion of Political Economy conference in Portugal in the September 2016. In the middle Lena Lavinas who gave a panel presentation
We researched these issues as a team of five: the principle investigator, Andrew Fischer; a postdoctoral researcher, Charmaine Ramos, and three PhD researchers, Ana Badillo Salgado, Emma Dadap Cantal, and Benedict Yiyugsah. We conducted our research through a mixed-method within-country and cross-country comparative case study approach applied to seven developing country cases (Cambodia and Philippines in Asia, Ethiopia, Ghana and Zambia in Africa, and Ecuador and Paraguay in Latin America). Our approach combined exploratory quantitative research with qualitative process tracing based on elite interviews and documentary research.
In our current, final stage of the project, we have been developing a set of propositions, which we have been writing up in a number of papers and one co-authored book.
With respect to the monetary dilemmas of transferring external resources to domestic expenditures such as social protection, we definitely observed much convolution in this regard. We also found evidence in some of the cases that it was exacerbating mistrust in donor-recipient relations due to concerns of transparency and accountability (with regard to donors, not necessarily to citizens) or else due to fundamentally different perspectives regarding the role of aid. This helps to explain the strong emphasis of surveillance and monitoring of CT programmes by donors in countries where donors play a strong role in these programmes (especially in Africa), arguably because donors can only guarantee that governments honour their domestic spending agreements through administrative methods of surveillance. That being said, we also found such forms of monitoring in countries where aid financing is relatively unimportant, such as in the Philippines, indicating other reasons for such measures, as discussed further below.
‘social protection is not directly financed but instead is mentioned as a condition or agreement....Such processes notably tie social protection programmes to the logic of conditionalities and IFI reform programmes.'
So-called social protection financing has also been generally disbursed through programme or policy loans and grants, rather than project loans/grants, except in our African cases and in Cambodia, where donors were directly financing programmes (and in the case of Cambodia, one donor was even running the largest CT programme as a subcontracted arrangement with USAID). There can often be a lot of ambiguity in this regard as well. In the Philippines, we actually spent a considerable amount of time trying to clarify whether a certain social protection associated loan from an IFI was a project loan (the IFI technical staff person understood that it was), or a programme loan (as the government understood). The loan documents were also ambiguous, with wording that could imply either type of loan. It was only a senior staff person in the finance ministry who was able to clarify that, in earlier negotiations, the IFI and the government had agreed to keep this ambiguity, in part to satisfy political sensitivities on both ends, but that the loan was effectively a programme loan.
Governments might prefer programme loans given that they can be freely used for budget support. However, we found that programme loans in our two Latin American cases to be unpopular given that they were associated with the legacy of structural adjustment programmes. The fact that social protection was associated with such loans was a sensitive political issue over which donors treaded lighted, particularly that borrowing to pay for social expenditures (or any current expenditure) was prohibited in both countries.
Andrew and Ana conducting interviews in Paraguay
In these cases, social protection is not directly financed but instead is mentioned as a condition or agreement, such as a commitment to increase government spending in a cash transfer programme, possibly by the amount of aid received from the donor, or possibly not even by a specified amount. Indeed, as long as spending increased within a specific earmarked programme, there did not appear to be any concern among major donors, including IFIs such as the World Bank or Asian Development Bank, whether this increase represented a net addition to overall social expenditure, or whether it was simply being taken away from other programmes, as a substitution with no aggregate net gains in social expenditures. In this manner, commitments to increased spending in cash transfers could be achieved while still adhering to broader caps in overall government spending. This is an especially sensitive issue in light of the constant lobbying by IFIs and donors to fund cash transfers with resources freed up by ending subsidy programmes, which governments in most of our cases have been resisting.
EU grant funding for social protection, for instance, operates in a similar manner, whereby a demonstrated commitment to a cash transfer programme is one of many triggers in a grant agreement, which unlocks a particular tranche of general budget support funding. In the case of one EU grant that we studied in Paraguay, the amount unlocked was one million euros per tranche, for a total of 24 tranches, which is arguably a very small sum relative to the onerous bureaucratic processes of negotiation and monitoring involved in unlocking it. In this case, EU staff were clear that the purpose of such funding was to influence policy, not necessarily to fill funding gaps in any significant way.
Such processes notably tie social protection programmes to the logic of conditionalities and IFI reform programmes. This point is mostly overlooked in the scholarship on external influence, given that such conditionality is mostly implicit. For instance, in the case of the EU funding mentioned above, one of the triggers is a commitment to macroeconomic stability, which is evaluated by the government committing to IMF Article Four consultations. In this sense, a structural adjustment programme might not be explicitly mentioned in an EU (or World Bank, etc.) grant (or loan) agreement, but it can be implied and crucially reinforced by such agreements, with donors following the IMF’s lead.
Fundamental dissonance in perspectives on the role of aid financing was especially revealed by our research in Ethiopia. The donors we interviewed there viewed aid as primarily an instrument of influencing recipient government policy through technical assistance, and the redistributive role of aid was taken for granted. Their main concern was sustainability, which they understood as the government gradually taking over the funding of its Productive Safety Nets Programme (PSNP) as the donors were scaling down and withdrawing. Interviews with the finance ministry, however, revealed a concern with sustainability that was primarily aimed at sustaining foreign exchange inflows into the economy in order to support much broader development efforts, such as infrastructure construction, and industrial and agricultural policy, all of which had been generating very deep trade deficits. Moreover, given that the government had agreed with the IMF in 2017 to cease all foreign borrowing, grants from donors were one of the few sources of foreign exchange remaining. Yet, my suggestion to a donor staff person that this was the guiding concern for the government and that financing social protection through domestic resources was much less of a concern was greeted with derision. We also observed this dissonance in Cambodia, where the government had remained entirely reliant on grants as their only source of external funding and hence the government’s principle motive for going along with various donor-lobbied social protection programmes was to mobilise foreign exchange through grants.
With regard to the other focus of the research on the political economy of social policy and how it is influenced by such external pressures, we have come to the conclusion that the dominant analytical frame that distinguishes between external versus domestic factors is misguided and unhelpful. Instead, it is far more useful to understand political economy dynamics in terms of institutional polarisation, with externally-oriented incentive structures and networks of policy elites that are as much transnational as they are domestic actors. Indeed, such polarisation was evident in even those cases where aid-dependence was insignificant in financial terms, such as Ecuador, Paraguay and the Philippines, which helps to explain why patterns of influence in these countries were similar to the more aid-dependent cases. While this is commonly referred to in the policy diffusion literature as ideational influences or epistemic communities, our contention is that these concepts do not go far enough into the structuring of elite incentives that guides policy making through power and resources as well as ideas and communities.
From this broader theoretical perspective, we have been formulating three overarching observations that are common to all seven countries in various ways. One is the superficiality of CT programmes relative to local political economy dynamics. None of the CT programmes in any of our case countries – of the sort lobbied by international donors – were supported, demanded, or otherwise influenced by the conventional social and political forces identified in the political economy of social policy, be it labour unions, social movements, political parties, business groups, or other power brokers at local or national scales. The lack of popular mobilisation is particularly interesting given that such mobilisations are conventionally assumed in the social policy literature (dominated by a northern perspective) to be a condition for social policy expansion. Rather, the policy making related to the introduction and even implementation of CT programmes was quite insulated, occurring as top-down technocratic policy innovations.
'all of the governments in our study – even the poorest and most aid-dependent – were funding their own preferred social protection or related income-supporting programmes entirely from their own domestic fiscal resources'
This is not to deny the relevance of conventional political economy analysis in such contexts. Rather, as our second observation, the dominant social and politic forces and distributional conflicts in each case were simply not focused on CT programmes, but instead on other priorities deemed more relevant. This point is exemplified by our observations that all of the governments in our study – even the poorest and most aid-dependent – were funding their own preferred social protection or related income-supporting programmes entirely from their own domestic fiscal resources, even whilst the notional funding of targeted CTs with aid is justified in terms of fiscal constraints. Examples include food or fertiliser subsidy programmes in Africa and elsewhere, minimum wage legislation (as in Cambodia among textile workers), and universal pensions. The fact that these other programmes are generally much larger than the donor-endorsed cash transfer programmes belies the claim that external resources are required for the latter.
This reinforces the point made above that the motivations for receiving aid for donor-endorsed CT and related programmes are often other than a shortage of domestic fiscal resources. Instead, they are more likely related to the need to mobilise foreign exchange for broader development efforts or for balance of payments support. In this sense, the adoption of these programmes appears to have often been used strategically by at least parts of the recipient governments, as a means to superficially demonstrate acquiescence to donor demands or expectations. As noted above, if and when donors perceive these government strategies, it exacerbates their concerns with the government’s commitment and transparency.
Charmaine with members of the civil society group Social Watch
However, as our third overarching observation, even where CT programmes are introduced in a marginal manner and without strong government commitment and with superficial insertion into local political economy dynamics, they can nonetheless carry important institutional inertia. Moreover, these programmes have served as vehicles for the normalisation of a particular targeting modality across the wider social protection/policy system, thereby transforming the nature of these systems once inserted from the margins.
These observations highlight the importance of adjusting our theoretical and analytical frames in countries characterised by high degrees of external dominance and orientation, as well as strong degrees of polarisation and disarticulation between elites and the bulk of the population. The fact that cash transfer programmes appear to instil an institutional inertia of their own within such contexts, despite the lack of proximate factors conventionally assumed to be important for social policy expansion, also provides important insights into both the social policy and institutionalist literatures. The implications – particularly where relatively small influences from the margins might gradually catalyse a transformation of a system – are very important for understanding the nature of external pressure, the evolution of social policy systems, and the nature of institutional change more generally in such circumstances. The fact that such effects could be achieved with very small amounts of aid and within broader policy environments that are non-conducive for development highlights some of the systemic political as well as economic challenges facing global redistribution towards poorer countries, particularly with respect to supporting the development of comprehensive social policy systems in such countries, while at the same time respecting ownership over national processes of policy making.