Policy-making is rarely straightforward. No issue area stands alone, and multiple different goals and motivations may be relevant at any given time. Thus aid policy cannot be reduced simply to the expression of a ‘moral vision’, nor to any particular ‘national interest’. It is, rather, governed by the interaction of many different frames for aid. As I have argued throughout this study, this makes it akin to a Swiss army knife: one tool, many functions. Or, as Italian legislator Grassi put it in 1995: aid is best seen simply as an ‘instrument of which the large Western countries avail themselves today’. As such, even though Italians called into question one long-established frame for aid – economic self-interest – in the mid 1990s, Grassi argued that aid’s value as an instrument for other frames implied a need to ‘retain and safeguard our own development cooperation policy’. At the same time, and contrary to the note of resignation that comes through in the quotation from Lancaster’s valuable recent book on aid, this study has shown that it is in fact possible to develop a model that is both elegant and insightful, while remaining comparatively parsimonious.

This concluding chapter is divided into three parts. The first section briefly reviews the material covered in the preceding chapters. Next, I provide some anecdotal evidence to demonstrate the relevance of the model to other aspects of aid policy as well as to the aid policy of other donor states. Finally, I suggest how the model might be applied to the analysis of other policy areas in foreign as well as domestic politics, and discuss the challenges inherent in applying a framework that requires measuring how decision-makers think about a particular policy.

Aid levels of the OECD donor states have varied considerably over time, ranging from virtually nothing to well over 1 per cent of GNP. Interestingly, despite four decades of activity in standard-setting and monitoring of donor state performances, the OECD’s Development Assistance Committee shows no signs of producing convergence among individual donor states on the international norm of 0.7 per cent of GNP. Instead, aid levels in many donor states dropped throughout the 1990s, moving further away from the norm, before increasing again during the first decade of the twenty-first century, albeit without reaching the levels of the 1980s.

Explaining patterns in overall aid levels has proven difficult: neither changes in government composition nor the economic fortunes of the donor states correlate highly. The end of the Cold War may have had an impact, but if so, it was in the opposite direction of that expected: many observers initially predicted aid levels would rise as military expenditures declined. Some explained the declining levels by pointing to public disillusionment or to a waning political commitment, but arguments along those lines date back at least to the mid 1960s, making them poor explanations for trends several decades later. In this chapter, I argue that the impact of various explanatory factors depends crucially on the relative strength of different frames for aid. The data about the salience of these frames presented in Chapter 3 help us explain patterns in aid volume from the 1950s throughout the end of the twentieth century.

Development assistance can serve a wide range of possible goals. Which goals matter most will depend on the way policy-makers frame the issue. This chapter presents measures of the relative prominence of different frames for aid, illustrating variation across both time and space. Although numerous patterns are shared across all four countries studied, their national aid discourses display enduring differences, dating from the early years of development assistance and lasting until the present day.

In Belgium, the frames of obligation and economic self-interest have dominated the aid discourse. Economic self-interest has also been central to Italian discussions of aid policy, but in that country this frame has been accompanied by an interest in the reputational aspects of aid. In turn, reputation has been an important secondary frame in Norway as well. However, in that country it has been coupled with a dominant emphasis on humanitarianism. In the Netherlands, finally, the two most prominent frames in relative terms have been humanitarianism and power.

Preface

I have been interested in foreign aid for almost as long as I can remember. Growing up in the Netherlands, I read the free monthly magazine Sam Sam (Working Together), published by the Dutch government to educate children about development assistance and distributed to every student in the upper grades of elementary school. When I began to pay attention to national politics some years later, I took for granted that although specific features of the development aid programme were frequently debated in the legislature and in the media, the need to have a large and generous aid programme was rarely questioned. Indeed, it was not until I moved to the United States for college that I encountered arguments against foreign aid per se.

If Americans had a different view of the practice of development assistance policy, they also appeared to have a different view of the theory that might explain it. Rational choice theory, rather more popular in the United States than in Europe, seemed to suggest that government policies must always pursue material interests, which made aid policy – with its ostensibly altruistic goals – hard to explain. I began to wonder how aid policy could be perceived so differently in the Netherlands and the United States, even though the policy instruments were essentially the same; I also wondered whether these differing perceptions might explain how widely aid programmes varied across donor states; and I wondered how the logic of rationality might account for a policy that seemed to have so many different possible motivations.

1. Take up the White Man’s burden –

2. The savage wars of peace –

3. Fill full the mouth of Famine,

4. And bid the sickness cease.

5. – Rudyard Kipling, 1899

Why do countries give foreign aid? Some earnest idealists see aid as a modern form of Kipling’s ‘White Man’s burden’: a worthy, noble enterprise, aimed at lifting those worse off than ourselves out of poverty. More critical observers point to the poem’s call for expansionism and Western control, and condemn aid as simply a modern form of imperialism. Others, still, note that Kipling may have intended his poem as satire, criticizing foolish notions about both the value and the feasibility of assisting those allegedly in need of superior Western beliefs, skills and products; aid, they suggest, is a similarly misguided and often counterproductive policy. There may be kernels of truth in each of these characterizations, but even taken together they offer at best an incomplete picture of the multifaceted policy area that is contemporary foreign aid. Aid programmes can handle whatever policy-makers put their minds to, making them the foreign policy version of a multitool or Swiss army knife.

Every advanced industrialized nation has a foreign aid programme, and each of these programmes officially aims to foster the development of denizens of the poorest countries. The sums involved are considerable. In recent years, annual transfers have exceeded $100 billion, which translates to over $100 per donor state citizen per year. In some donor states the aid programme accounts for 5 per cent or more of the government budget. On the recipient side, official development assistance (ODA) accounts for a large share of international capital flowing into less developed countries (LDCs). Nevertheless, the factors shaping foreign aid remain ill-understood, more than half a century after Morgenthau first described the policy as ‘baffling’. Aid levels rise and fall without obvious causes. Explanatory factors that appear important in one case are insignificant in another. And case studies of different aid programmes frequently explain similar empirical patterns using incompatible models. One of the best recent studies on foreign aid frankly concludes that ‘There are too many interacting variables to justify a model that would be both parsimonious and insightful’.

Chapter 3 highlighted the striking differences, both across donor states and over time, in the framing of development assistance by decision-makers. This chapter looks into the sources of these differences, and analyses the factors that cause ideas about aid to change. In the next few chapters, I will argue that frames for aid are both prior to and shape aid policy; this makes it all the more important to provide an account of the sources of these frames. I do so by relating them to deeper beliefs about national identity and historical experiences. I shall argue that four broad factors affect the relative strength of different frames: national traditions and experiences, the international political and economic context, the efforts of governments as frame entrepreneurs and information providers, and experiences with past policy successes and failures. International norms, on the other hand, exert a weak influence at best.

The chapter is divided into four sections. First, I examine the connections between national identity and the initial discourse on aid that emerged in the 1950s. Since multiple policy frames may be compatible with a nation’s history and identity, it is especially important to examine the early debates on foreign aid, as that is where the initial set of relevant arguments will be identified and tested. The second section discusses in more detail the evolution of aid discourses in the four countries. This allows us to see how international influences as well as national experiences influenced these discourses over time. The third section takes a closer look at the impact of governments as frame entrepreneurs. I show that elites are far from passive participants in the policy discourse. Instead, they actively work to increase the prominence of the frames they consider most important, whereas others are downplayed as much as possible. In the fourth section, finally, I discuss the way governments may control the visibility of aid policy and the availability of information about successes and failures, in attempts to prevent or reinforce shifts in the relative salience of different frames.

This appendix lists the parliamentary aid debates coded in each country. The main source of the data for each year is the debate about that year’s development cooperation budget. Sometimes additional debates were coded in the same year, such as general debates on the overall development assistance programme of the government. Occasionally, additional debates for the preceding year were included in order to get a slightly larger sample of speaking turns. This was useful particularly for Belgium and Italy, where aid debates have been both less extensive and more legalistic than in the Netherlands or Norway. As a result, each data point is less a snapshot than an average impression of the discourse over a two to three year period.

Belgium

Belgian budgets are debated in parliamentary committee first, before being reviewed and voted on in a plenary session. Both the Senate and the Chamber of Representatives have to approve the budget in a plenary session, but one Committee does the bulk of the work, preparing a report for the full session. In even years, the committee of the Chamber has this duty; in odd years, it falls to the Senate. The locus of initial discussion tends to feature more in-depth policy debates. For consistency’s sake, however, only debates of the Chamber or its committees were coded. The Minister or State Secretary in charge usually opens the committee debates, then takes a series of questions by committee members, and (usually, but not always) answers these. The plenary discussion features a brief report by the committee chair, followed by statements on the budget and on aid policy in general by representatives from all of the major parties, together with some additional questions for the Minister or State Secretary.

What might Belgium’s Development Minister have meant when he argued that his country’s aid programme was ‘least bad’? Regular reviews of aid programmes produced by the Development Assistance Committee suggest that an important part of the answer lies in the overall administration of aid policy. When the DAC reviewed Belgium’s aid programme in 2001, it emphasized the implementation of ‘major legislative changes and sweeping administrative reforms’. Similarly, the central finding in its 2000 review of Italy’s policies was that ‘Considerable progress has been made in the management of Italian aid’.

Another important factor in evaluating an aid programme is the quality of the aid it disburses: the fewer strings are attached to particular aid flows, the higher the quality is generally deemed to be. This chapter analyses both the overall administration of the aid programmes in Belgium, Italy, the Netherlands and Norway, as well as their performance on two key indicators of aid quality: tied aid and multilateral aid. Among others, this will help us decide the accuracy of Minister Outers’ assessment of his country’s programme.

Data for the dependent variable was derived from the aid database of the OECD’s Development Assistance Committee (DAC), both in CD-ROM format and online (www.oecd.org/dac). As discussed in Chapter 7, the aid measure selected was net disbursements. Occasionally countries repay earlier development loans, which makes it appear as though they receive negative aid. However, we are concerned with a recipient’s share of a donor’s total aid in a given year. Since negative aid is equivalent to a zero share, negative values are truncated to 0. For every recipient-year, I calculated the aid share of that recipient from each of our four case study countries, as well as its aid share from the DAC as a whole. Often, a few countries get most of the aid, while a much larger number gets comparatively little. In order to generate a more meaningful ranking, therefore, I take the log of these aid shares.

Data for the independent variables was drawn from a number of different databases. Basic country indicators – GDP, area and population, as well as literacy, life expectancy and mortality rates – are from the World Bank’s World Development Indicators CD-ROM (2004 edition). Trade data were taken from Gleditsch’s trade and GDP dataset. Data on political similarity is provided by Gartzke’s Affinity of Nations dataset, which provides an indicator of voting agreement at the United Nations. Data on the quality of democracy (or autocracy) in a country were taken from the Polity IV dataset, and information about political and civil rights came from the Freedom House dataset.

This appendix gives sample quotations for each of the forty-four different arguments for aid that were coded. Arguments are listed under the general frame to which they were allocated. Note that certain arguments were allocated to several frames, as described in Chapter 3. However, each argument is listed only once here. For each quotation, I list the legislator, his or her country, and the date and location in the parliamentary record.

In fact, Haugland wanted to include development aid as part of the defence budget, ‘as a demonstration of our interest in peace and of our way of practising international politics’ (ibid.). Interestingly, in this second quotation, the reputation frame takes the upper hand.