1-s2.0-S0305750X1500011X-main (1)

8/17/2019 1-s2.0-S0305750X1500011X-main (1)

1/20

Diverging Visions on Political Conditionality: The Role

of Domestic Politics and International Socialization

in French and British Aid

DAMIANO dE FELICE*

London School of Economics and Political Science, United Kingdom

Summary. — This article shows that and explains why the United Kingdom has internalized political conditionality to a larger extentthan France. The assessment part is based on the analysis of policy documents, international agreements, and two “hard” cases(Mozambique and Zimbabwe). Variation between the two countries is explained by the existence of clearer lines of accountability forBritish aid decision-makers, stricter scrutiny by British media, and stronger social pressure from Nordic donors. Evidence does notsupport the explanatory power of material interests, party politics, level of parliamentary control, and socialization processes withinthe Commonwealth (versus the International Organization of La Francophonie). 2015 Elsevier Ltd. All rights reserved.

Key words — aid, political conditionality, human rights, Europe, United Kingdom, France

1. INTRODUCTION

After the end of the Cold War, France and the United King-dom (UK) were among the first donors to signal parallel pol-icy changes in favor of political conditionality, that is, thenorm by which the allocation and disbursement of bilateraldevelopment assistance should depend on respect for humanrights and democratic principles by recipient governments.On June 6, 1990, the British Secretary of State for Foreignand Commonwealth Affairs Douglas Hurd stated that “gov-

ernments which persist with repressive policies, corrupt man-agement, wasteful and discredited economic systems shouldnot expect us to support their folly with scarce aid resources”(quoted in Robinson, 1993a, p. 88). On June 20, 1990, theFrench President, François Mitterand (1990a), concluded theFranco-African Summit at La Baule by affirming that “Francewill bind all its [aid] contributions to the efforts that will bemade to move towards more freedom”.

This novelty was not to be an ephemeral moment for thedonor community; political conditionality was subsequentlyhailed as the “core” of the post-Cold War international aidregime (Gibbon, 1993, p. 36), a “strongly coercive” trend(Baylies, 1995, p. 328), and a “new standard” in developmentcooperation (Arts, 2000, p. 1). In 2007, Carey (2007, p. 460)

commented that international norms still “

legitimize and evenencourage such donor behavior”.Against this background, it is puzzling to discover that

today Paris and London hold strikingly different views withrespect to political conditionality. In 2011, during the consul-tation on the Green Paper “The future of European Union(EU) budget support to third countries”, the EuropeanCommission (2011) explicitly asked: “should budget supportprograms make more use of political governance condition-ality?”. 1 The French government argued that “budget supportcannot be conceived as an instrument to promote values andpolicy objectives, except if you want to divert the purpose”,that is, “to support the national strategies to fight againstpoverty, elaborated by partner countries” (French Ministryof Foreign, 2011b). In contrast, the UK stressed the need to

“raise political governance issues through its dialog on budgetsupport with partner governments—with the clear focus oncommitment to fundamental values of human rights, democ-racy and the rule of law” (European Scrutiny Committee,2011).

This article shows that and explains why the UK has inter-nalized political conditionality to a larger extent than France.Assessment of internalization is based on a three-fold measure-ment framework (policy statements, legal texts and statebehavior). The analysis of around 70 strategy papers and

international agreements is corroborated by the examinationof two “hard” cases (Mozambique and Zimbabwe). Explana-tion of internalization combines a Most Similar SystemsDesign with process-tracing. In order to test the relativeexplanatory power of material interests, domestic politics,international socialization pressures, and organizational cul-tures, information from primary sources (such as independentnewspapers’ articles, government evaluation reports anddiplomatic cables) was triangulated with semi-structured inter-views with more than 100 officials who are working or haveworked for French and British institutions, and aid practition-ers who are working or have worked for other donors or fordevelopment/human rights NGOs. 2

The article offers two main contributions to the existing lit-

erature on aid, human rights, and democracy. First, it enlarges

* Earlier versions of this article were presented at the ECPR Joint Sessions

of Workshops in Mainz, Germany, in March 2013, and at the EISA Pan-

European Conference on International Relations in Warsaw, Poland, in

September 2013. I am extremely grateful to the participants in these events

for their helpful comments—in particular to Karen Del Biondo, Jonathan

Fisher, Christine Hackenesch, Wil Hout, Svea Koch and Clara Portela.

Advice and support from Sebastian Dellepiane, Jörg Faust and Nadia

Molenaers have been invaluable, and I would like to thank them sincerely

too, together with Karen E. Smith, Chris Alden, Roberto Roccu, Bene-

detta Voltolini and three anonymous reviewers. Finally, I would like to

thank the numerous British and French officials who responded promptly

and thoroughly to my persistent questions. All errors are my own.

World Development Vol. 75, pp. 26–45, 20150305-750X/ 2015 Elsevier Ltd. All rights reserved.

www.elsevier.com/locate/worlddevhttp://dx.doi.org/10.1016/j.worlddev.2015.01.010

26

http://-/?-http://-/?-http://-/?-http://-/?-http://dx.doi.org/10.1016/j.worlddev.2015.01.010http://crossmark.crossref.org/dialog/?doi=10.1016/j.worlddev.2015.01.010&domain=pdfhttp://dx.doi.org/10.1016/j.worlddev.2015.01.010http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-

8/17/2019 1-s2.0-S0305750X1500011X-main (1)

2/20

the scope of investigation to the analysis of internationalagreements and policy documents. Past research on politicalconditionality has given almost exclusive attention to thebehavioral aspect of this norm. Quantitative researchers haveexplored whether human rights records have influenceddonors’ decisions on (a) who their recipient governmentsshould be and (b) how much aid these governments should

receive (for instance, on US aid allocation see Abrams &Lewis, 1993; Apodaca & Stohl, 1999; Cingranelli &Pasquarello, 1985; Poe & Meernik, 1995; Schoultz, 1981).Qualitative studies have concentrated on specific cases whendonors applied (or did not apply) aid sanctions (Renard &Reyntjens, 1995; Schulte Nordholt, 1995; Stokke, 1995;Waller, 1995). In contrast, comparative research on politicalconditionality and international agreements is non-existent,and only a couple of researchers have looked at the policyagendas of different donors on a comparative basis (Barratt,2008; Crawford, 2001).

Second, past studies have shown that some donors applypolitical conditionality more consistently than others. Forinstance, numerous statistical analyses have contrasted theinfluence of human rights concerns on aid allocation by differ-

ent donors (Berthélemy, 2006; Carey, 2007; Clist, 2011; Dollar& Levin, 2006; Hoeffler & Outram, 2011; Neumayer, 2003;Younas, 2008; Zanger, 2000), and qualitative researchers com-pared donors’ responses to specific cases of human rightsviolations and democracy setbacks (Crawford, 2001, pp.163–181). However, while variation was established, no effortwas made to explain it. This article builds on the only studythat has explored why (not only that) donors internalize poli-tical conditionality to different degrees (Cumming, 2001), butexpands its analysis to 2012 and tests new relevant hypotheses(for instance, Cumming overlooked the potential influence of organizational cultures and international social pressures fromlike-minded donors and international organizations).

The article is structured as follows. The second section jus-

tifies the selection of France and the UK. The third sectioncompares the two donors on the basis of (1) endorsement of political conditionality in policy documents, (2) inclusion of human rights clauses in international agreements, and (3) will-ingness to adopt aid sanctions in response to human rightsviolations or democratic setbacks in recipient countries. Thefourth section embraces “analytic eclecticism” and tests fourdifferent hypotheses that can explain variation across theChannel. The conclusion summarizes the main findings of the article and recommends a potential avenue for futureresearch.

2. SELECTION OF FRANCE AND THE UK

The Most Similar Systems Design (or Mill’s Method of Dif-ference) suggests comparing cases that are as similar as possi-ble, except with regard to the dependent variable. Theambition is to keep constant the highest possible number of independent variables (Anckar, 2008; Yin, 2009, pp. 64–67).Bearing this in mind, the selection of donors for this researchwas narrowed down through a two-step process.

To begin with, the universe of potential cases was restrictedto EU Member States. The EU as a whole is one of the mostenthusiastic devotees of political conditionality. Since 1990,the Commission and EU Member States have consistentlyinserted a human rights clause in all their developmentagreements with third countries (Bartels, 2004; Horng,2003). In addition, EU aid has been suspended, redirected,

or withdrawn in 56 cases, an average of three times per year(see Table 1).

Strong internalization of political conditionality by EUinstitutions encourages a similar position to be held by Franceand the UK for two main reasons. On the one hand, aid sanc-tions are usually decided through unanimous agreement with-in the Council. This means that when EU aid is suspended,

reduced or redirected, Member States have agreed on theappropriateness of these measures. Second, the academic lit-erature on “Europeanization” has persuasively demonstratedthe harmonizing power of EU institutions with respect tomember States’ politics and policies (Featherstone &Radaelli, 2003; Ladrech, 2010), even in areas complementaryto foreign aid, including foreign policy (Tonra, 2001; Wong& Hill, 2011).

Within the EU, the article selected France and the UKbecause of their similarities as middle-ranking powers, nucle-ar-weapon States, permanent members of the United NationsSecurity Council, former colonial empires and human rights“homelands”. Moreover, in 1998 the Heads of State andGovernment of the two countries met in Saint-Malo, andagreed on a declaration putting an end to Anglo-French rival-

ry in Africa (Chafer & Cumming, 2010). This initiative “estab-lished the basis for . . . harmonizing policies and the overallapproach” toward the continent (Jones-Parry, 2011, p. x),and led to initiatives like cooperation between Heads of Mis-sion in individual countries and informal dialog within Euro-pean fora (including the Africa Working Group and theGeneral Affairs and External Relations Council, that is, wherepolitical conditionality is discussed) (Cumming, 2011, p. 59).As Africa represents by far the most recurrent target of Euro-pean aid sanctions, Saint-Malo should have led to enhancedunderstanding between both countries.

3. ANALYSIS OF VARIATION

Assessing the significance of human rights norms in the for-eign policy of a specific country is a difficult exercise. As high-lighted by Brysk, “some countries claim more than theyproduce . . . in order to gain international reputation. Con-versely, principled promoters often under-assess their ownefforts, discounting modest gains relative to unfulfilled aspira-tions” (2009, pp. 19–20). Past literature in IR offers little help.For instance, Cortell and Davis commented that “scholarsrepeatedly conclude that domestic salience is crucial to manycases of states compliance with international norms, but theyrarely provide definitions or operational measures for the con-cept and, instead, merely assert that the norm in question wassalient” (2000, p. 67).

This article argues that, if we accept the definition of a norm

as a “standard of appropriate behavior for actors with a givenidentity” (Finnemore & Sikkink, 1998, p. 819), norm internal-ization has three fundamental indicators when applied todonor countries:

1. Policy documents. Since a norm is a standard of appropriate behavior, and appropriateness is communicativelyshared, a salient norm is referred to by the State to justifyits behavior.2. Legal texts. Since a norm is a standard of appropriatebehavior, a salient norm is institutionalized into the docu-ments that regulate development cooperation.3. State actions. Since a norm is a standard of appropriatebehavior, a salient norm is acted upon by the State (for asimilar three-fold measure, see Farrell, 2001, p. 79). 3

DIVERGING VISIONS ON POLITICAL CONDITIONALITY 27

http://-/?-http://-/?-http://-/?-http://-/?-

8/17/2019 1-s2.0-S0305750X1500011X-main (1)

3/20

(a) Analysis of variation in policy documents and legal texts

At the discursive level, internalization of political condition-ality is deep when ministerial speeches and strategy papershabitually refer to respect for human rights as a pre-conditionfor aid allocation and disbursement. Internalization is moder-ate in case of doubts about the effectiveness of aid sanctions.For instance, donors may worry about the negative conse-quences of reducing resources in favor of the poor population,or favor the strategy to strengthen economic and political linkswith repressive regimes thus engendering a process of internalchange (Lavin, 1996; Smith, 1998). Internalization is shallow

when non-compliance is explicitly justified by reference tothe inappropriateness of the norm. For example, new donorsreject political conditionality as an act of neo-imperialism(Barya, 1993).

As regards legal texts, international agreements with recipi-ent countries offer the opportunity to establish reciprocal obli-gations and clarify the conditions for the implementation of development projects. Deep internalization of political condi-tionality pushes donors to systematically include a humanrights clause which specifies that lack of respect for humanrights and fundamental democratic principles can lead to sus-pension or cancelation of aid programs. As an example, the

Table 1. European Union aid sanctions during 1991–2013

Aid sanctions decided by the Council after consultations under a human rights clause

Central African Republic 2003Union of the Comoros 2000Republic of Côte d’Ivoire 2000 and 2001Republic of Fiji 2001 and 2007Republic of Guinea 2005 and 2009

Republic of Guinea-Bissau 2011Republic of Haiti 2001Republic of Liberia 2002 and 2003Republic of Madagascar 2010Islamic Republic of Mauritania 2006 and 2009Republic of Niger 1996, 1999, 2010Republic of Togo 1998 and 2004Republic of Zimbabwe 2002

Aid sanctions decided by the Council

Republic of Belarus 1997

Aid sanctions decided by the Commission

Republic of the Union of Myanmar 1991Republic of Burundi 1993 and 1997Kingdom of Cambodia 1997

Central African Republic 1996Federal Islamic Republic of the Comoros 1995Republic of the Congo 1997Republic of Croatia 1995Republic of Djibouti 1991Democratic Republic of the Congo 1992Republic of Equatorial Guinea 1992Republic of The Gambia 1994Republic of Guatemala 1993Republic of Haiti 1991 and 1997Republic of Honduras 2009Republic of Kenya 1991Republic of Malawi 1992Republic of Mali 2012Republic of Mozambique 2009

Republic of Nicaragua 2009Federal Republic of Nigeria 1993 and 1995Republic of Peru 1992 and 2000Republic of Rwanda 1994Russian Federation 2000Republic of Serbia 1998Republic of Sierra Leone 1997Syrian Arab Republic 2011Republic of Tajikistan 1998Togolese Republic 1992Republic of Uzbekistan 2005Zanzibar 1997

Sources: Bakari (2001), Bartels (2008), Bradley (2005), Brandtner & Rosas (1998), Brown (2005), Crawford (2001), da Câmara & Ferreira (2001),European Commission (2000), Fierro (2003), Hayman (2011), Hazelzet (2001), Ihonvbere (1997), Kreutz (2005), Laakso et al. (2007), Portela (2010),Saltnes (2013), Smith (2013), Uvin (1993) and Zimelis (2011).

28 WORLD DEVELOPMENT

8/17/2019 1-s2.0-S0305750X1500011X-main (1)

4/20

standard format for program support agreements signed bythe Government of Denmark plainly affirms that “respectfor human rights, democratic principles, including free andfair elections . . . constitute essential elements” of the instru-ment, and that in case of “violation of the essential elements. . . the Government of Denmark reserves the right to suspendwith immediate effect further disbursements to the implement-

ing partners”

(Government of Denmark, 2010).

(i) FranceThe discourse by Mitterand at La Baule on June 20, 1991

seems to indicate strong support of political conditionality.However, its significance should not be overestimated. At apress conference organized the day after the event, Mitterandclarified that “each country should set the terms and pace of its own reform” (Mitterand, 1990b). At the 1992 Franco-African Summit, the French Prime Minister Pierre Beregovoyprivately explained that among potentially conflicting goals,security ranked first, development should be pursued secondand only then could democratization be tackled (Renou,2002, p. 18). In 1993, the French Minister of Foreign Affairs,Roland Daums, stated that “economic reform should have

priority over democratization” (Sadie, 2002, p. 60).The 2006 strategy paper “Governance Strategy for French

Development Assistance”, approved by the InternationalInterministerial Cooperation and Development Committee(CICID), offers compelling evidence of disinclination againstpolitical conditionality. French policy-makers explicitlyacknowledge deep internalization of the norm by the UK:“other bilateral agencies have broadened their concerns toencompass a democratic agenda promoting a sound internalpolitical base as a prerequisite for more successful develop-ment programs. These agencies place great emphasis on gover-nance issues, as can be seen in the latest white paper onpoverty and governance by Britain’s Department for Interna-tional Development” (French Ministry of Foreign, 2006, p. 3).

In contrast, conditionality is expressly ruled out in Frenchdevelopment cooperation. The basic idea is that a “turnkey”human rights and democratic development model becomescontradictory if it means that local choices are pre-empted.As such, the quality of cooperation “should not be measuredso much by its ability to lay down universal standardsmanipulated in the abstract through conditionality as it shouldbe assessed by its ability to provide each partner with specificexperience and expertise to enable them to develop their ownpolicies” (French Ministry of Foreign & European Affairs,2006, p. 16). 4

This position was subsequently confirmed by numerous pol-icy documents. For instance, French budgetary aid is dis-pensed on the basis of a “Doctrine for the use of generalbudget support in foreign States” adopted in February 2007.

Recipient countries should meet three conditions in order tobe eligible for general budget support: a sound and sustainablemacroeconomic policy, a growth and poverty-reduction strat-egy in line with the Millennium Development Goals (MDGs)and a favorable assessment of the public financial manage-ment system (CICID, 2007). No human rights pre-requisiteis included.

In 2011, the French government adopted its first long-termpolicy strategy: “Development Cooperation: A FrenchVision”. The document reiterated that France decided to“move from a development assistance policy which impliesan asymmetrical donor–beneficiary relationship over to a pol-icy of cooperation with developing countries” (FrenchMinistry of Foreign, 2011a, p. 6). This entails that “coop-eration policy . . . should not use an unequal balance of power

to impose a particular form of governance on its partners”.Support for good governance “comes up against the limits,encountered by any outside party, of non- interference in thepolitical and social balance of a sovereign country” (FrenchMinistry of Foreign & European Affairs, 2011a, p. 33).

As far as development agreements are concerned, Francemainly concludes three types of legal instruments with recipi-

ent countries: Friendship and Cooperation Treaties, Partnership Framework Documents, and Debt Reduction and Development Contracts.

One of the most detailed Friendship and CooperationTreaty between France and a developing country was signedwith Afghanistan in January 2012. This instrument, whichcovers development cooperation, mentions human rights inthe preamble with the Parties concluding the agreement by“reaffirming their commitment to the principles of nationalsovereignty, democracy, human rights and equality betweenmen and women enrolled in their Constitution”. However, thisreference is evidently weak, not only because of its marginalplace (the preamble has no “hard law” effects), but also forthe juxtaposition of the concept of national sovereignty.

Democratic governance and rule of law are also considered,but only as fields to be supported, not as conditions of theagreement (Article 8). The lack of any political conditionalityis confirmed by the scant termination clause: “this treaty maybe terminated in writing by either of Parties with threemonths’ notice through diplomatic channels” (Article 13).

In 2004, the CICID institutionalized the preparation of Partnership Framework Documents (Documents Cadre deParté nariat, DCP) in order to set the priorities of French aidin its priority countries. DCPs are signed by both Franceand the recipient government for a five-year term, and aredrafted in accordance with the partner country’s developmentstrategy. French DCPs never include human rights clauses.When reference to human rights is present, this is only to men-

tion projects related to the enhancement of the rule of law. Forinstance, the DCP signed with Ethiopia in 2006 includes a sec-tion (4.2.1.1) titled “Support to the judiciary and lawreforms”, under which France committed to “contribute tothe following activities to strengthening of democracy, the pri-vate sector and access to justice: . . . to support the traininginstitute for judges and prosecutors”.

Debt Reduction and Development Contracts (C2Ds) areFrance’s bilateral instrument for reducing debt incurred aspart of development assistance. They are signed and imple-mented after the completion point of the multilateral HeavilyIndebted Poor Countries initiative and are added to the debtcancelations granted through the Paris Club. Since 2000,numerous countries under human rights scrutiny have signedC2Ds, including Uganda (2002), Mauritania (2003 and

2006), Madagascar (2005) and Cameroon (2006). None of the-se contracts and none of the relevant reports submitted to theNational Assembly have ever mentioned human rights.

(ii) United KingdomSince the end of the Cold War, British governments from

both sides of the political spectrum have recurrently endorsedpolitical conditionality. In 1997, the newly elected Labourgovernment created the Department of International Develop-ment (DfID). The first DfID white paper expressly requestedrecipient governments to “pursue policies which promoteresponsive and accountable government, recognizing that gov-ernments have obligations to all their people [and] promotethe enjoyment of civil, cultural, economic, political and socialrights” (DfID, 1997, p. 39). In 2005, DfID, the Treasury and


http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-

8/17/2019 1-s2.0-S0305750X1500011X-main (1)

5/20

the Foreign and Commonwealth Office (FCO) jointly adopteda document titled “Partnerships for poverty reduction:rethinking conditionality”, and set out the circumstances inwhich the UK would consider modifying existing aid commit-ments. The three ministries argued that “an effective aid part-nership should be based on a shared commitment to threeobjectives:

1. reducing poverty and achieving the MDGs;2. respecting human rights and other international obliga-tions; and3. strengthening financial management and accountability”(DfID, 2005, pp. 1, 8).

The option of “reducing or interrupting aid” has to be con-sidered when, among other conditions, “countries are in sig-nificant violation of human rights or other internationalobligations” (DfID, 2005, pp. 3, 9).

Since 2005, the conditionality policy has been continuouslyreaffirmed (see DfID, 2008, 2009). Importantly, this approachhas also been confirmed by the most recent Coalition govern-ment. In July 2011, DfID published a “Technical Note onImplementing DfID’s strengthened approach to budget sup-port”. In considering whether to give budget support, the

UK “will continue to assess governments against the threecommitments”, and it will place even “more emphasis ondomestic accountability” (DfID, 2011, p. 1).

Besides embracing political conditionality in policy docu-ments, the UK has consistently inserted a reference to theexistence of shared commitments to respect human rightsin its international agreements with developing countries(including overarching bilateral partnership arrangementsand joint minutes of aid talks). For instance, the UK havesigned several Development Partnerships Agreements(DPAs) with developing countries, among which Afghani-stan (2005), Vietnam (2006), Zambia (2006), Yemen (2007)and Uganda (2007). Development Partnership Arrangements(DPAs) are international agreements that set out the annual

volume of aid that DfID expects to allocate to a partnercountry over a ten-year period. All of these DPAs includea human rights clause within the “Basis of the Partnership”,which reads as follows: “Our development partnership in[Country X] is based on commitment to the following threeobjectives: (i) Reducing poverty and achieving the MDGs in[Country X]; (ii) Respecting human rights and other rele-vant international obligations; (iii) Strengthening financialmanagement and accountability and reducing the risk of funds being misused through weak administration or cor-ruption”.

Unlike DPAs, Bilateral Memoranda of Understanding withdeveloping countries do not feature a standardized formulafor the human rights clause. However, the large majority isbased on a mutual pledge to respect human rights. For

instance, in 2006 the Governments of the UK and Rwanda jointly committed themselves to, inter alia, “the promotionand protection of the full range of human rights of all Rwan-dans” (Article 6). Article 8 and 9 expressly regulate the termi-nation of the agreement in case of alleged human rightsviolations: “[s]hould either Government believe that the otherhas taken or plans to take action incompatible with a commit-ment within the MoU, then the parties will meet to formallyrecord the concern, draw attention to the potential implica-tions of a failure to honor a commitment, and discuss whatremedial action might be taken. If after substantive discus-sions either party continues to have significant concerns, thenthe UK’s program of assistance to the Government of Rwanda may be reconfigured or, in extreme circumstances,terminated”.

(b) Analysis of variation in behavior

History offers innumerable examples of divergence betweenwords and deeds. Following the argument that “the prefer-ences of policy-makers are revealed by the choices govern-ments make about resource allocation” (Barratt, 2008, p.208), many researchers have used econometric techniques to

investigate the influence of potential recipients’ human rightsperformance on aid allocation by bilateral donors, includingFrance and the United Kingdom. Table 2 offers a comprehen-sive overview of their results.

As can be seen, there is almost unanimous consensus thathuman rights matter less in France than in the UK.

These findings offer a useful starting point to assess theapplication of political conditionality. However, statisticalanalyses rely upon imperfect measures of aid sanctions. 5 Thismeans that “in an area as complex as aid, it is very possiblethat considerations can have important effects that are noteasily captured in aggregate analyses . . . Perhaps the mostdirect way of looking at the nuanced ways that a considerationsuch as human rights can have an effect is by looking at indi-vidual decisions themselves” (Barratt, 2008, p. 116; see also

Crawford, 2001, p. 8). This is particularly true also becausequalitative analysis provides important insights into hownorms shape both “events” (donors may offer different reasonsto suspend aid, such as human rights abuses or macro-eco-nomic imbalances) and “non-events” (only some donors may

justify the non-application of the norm, thus implicitlyacknowledging its salience) (Checkel, 1999, p. 87).Kratochwil and Ruggie (1986, pp. 766–769)) argued longago that norms serve as reasons for action rather than imme-diate causes, and that any explanation that seeks to invokenorms must do more than point to the existence of normson the one hand, and norm-consistent behavior on the other.

Past qualitative research strongly suggests that France hasbeen less willing than the UK to apply political conditionality.

Uvin found that, in comparison with other major donors, “

theFrench policy toward political conditionality is much moremodest . . . Generally spoken, its position continues to beone of silent support for the prevailing regimes in its formercolonies” (Uvin, 1993, p. 66). Cumming concluded that inthe 1990s there was “a radical shift with the introduction of political conditionality, but this has gradually been watereddown by France, and, to a lesser extent, the UK”(Cumming, 2001, p. 340).

Space constraints preclude for a comprehensive analysis of numerous qualitative studies. Since past research has alreadyoffered preliminary support to the thesis of the article, theremaining part of this section limits itself to offer some confir-matory evidence from the brief analysis of two additional cas-es. Research design rules suggest selecting the hardest possible

cases where one would expect confirmation of the startinghypothesis that the UK is more willing than France to applyaid sanctions. These cases offer the strongest possible evidencethat the hypothesis is supported by empirical evidence(Bennett, 2004; King, Keohane, & Verba, 1994).

First, the article focuses on Mozambique, a recipient coun-try where both France and the UK are considered to be “key”donors (Manning & Malbrough, 2010, p. 149). The similarityof the donor–recipient relationships makes Mozambique anexceptional case where Paris and London can be directly con-trasted one against the other. In addition, Mozambique is fre-quently cited as a best case of donor harmonization (Brütsch,2014, p. 219). High levels of donor coordination enhance theexpectation that the two donors should have behaved insimilar ways.


http://-/?-http://-/?-http://-/?-http://-/?-

8/17/2019 1-s2.0-S0305750X1500011X-main (1)

6/20

8/17/2019 1-s2.0-S0305750X1500011X-main (1)

7/20

Second, the article focuses on a recipient country with a spe-cial relationship with the UK. Such a case represents a “hard”test because the hypothesis is that the UK has internalizedpolitical conditionality to a larger degree than France. If theUK applies political conditionality in a case where it hasstrong interests to defend, the adoption of the norm can beconsidered particularly strong. If France is against the applica-

tion of political conditionality in a case where it has less press-ing objectives to pursue, the adoption of the norm can beconsidered particularly weak. Between Fiji and Zimbabwe— the two former British colonies which have been subject toEU aid sanctions under the human rights clause—the articleconcentrates on Zimbabwe because of the larger size of bothBritish and French aid programs in this country.

(a) Mozambique 6

A couple of months before the 2009 Mozambican elections,a new political party, the Democratic Movement of Mozam-bique, was barred from running in nine of the thirteen parlia-mentary regions (Astill-Brown & Weimer, 2010, p. 8;Manning, 2010, p. 156). In addition, on the day of the elec-

tions, independent observers documented widespread irregula-rities, including ballot stuffing and tabulation frauds (Hanlon,2010, p. 92; Lloyd, 2012, pp. 2–5). In response to these events(but also as a consequence of wider economic and politicalconcerns), in December 2009, the Group of 19 budget supportdonors (G19) went “on strike”, freezing all contributions for2010 (“Aid Strike in Maputo,” 2010).

Even though all G19 countries signed the letter thatinformed the Mozambican government of the donor strike,heated discussions took place on the appropriate course of action. Foreign diplomats report that the hardliners includedthe Nordics, Netherlands, UK, Ireland, Switzerland, andCanada. On the other end of the spectrum were Spain, Italy,France, and Portugal. 7

The most important news agency in Maputo confirmedFrance’s adoption of a soft stance: “some of the smaller bud-get support contributors (such as Portugal, Italy and France)have made it clear that they are not in favor of withholdingfunds” (“Government & Donors Claim Consensus,” 2010).Local commentators also interpreted a trip of the French Sec-retary of State for Foreign Trade, Anne-Marie Idrac, toMaputo at the beginning of March 2010 as “an indication thatFrance, like Portugal and Italy, does not regard the donorstrike as a useful tactic” (“No Indication That Donors WillEnd Support,” 2010).

With respect to the UK, the willingness to apply aid sanc-tions is confirmed by a diplomatic cable from the US Ambas-sador, who reported that “the UK High Commissioner, alongwith the DFID representative, were firm in supporting the

immediate declaration of a breach, an announcement of a real-location of some amount of budget support to other projects,and the commencement of negotiations with the governmentof Mozambique to reach an agreement on steps which mustbe taken to restore the totality of budget support”(Wikileaks, 2009, para. 4). Interestingly, for the UK the strikedid not mean only the temporary suspension of disbursements.At the ceremony when aid commitments were eventually deliv-ered, the G19 announced that several donors could haveincreased their budget support but chose not to; newspaperssingled out Britain and the Netherlands (“No BudgetSupport Received So Far This Year,” 2010). In June 2014 Bri-tain even announced that DfID would not disburse generalbudget support to Mozambique any more (“Aid RowEscalates,” 2014). There would instead be new forms of aid,

tied to indicators on good governance (“Mozambique: UKEnds Direct Budget Support to Mozambique,” 2014).

(b) Zimbabwe

In 2000, Robert Mugabe proposed a constitutional referen-dum to grant himself the right to stand as President for another

two terms and permit the confiscation of white-owned landwithout compensation (Sachikonye, 2002; Scarnecchia, 2006).The failure of the referendum represented a watershed momentfor future human rights violations and electoral irregularities inthe country. Since then, Mugabe fastened the illegal processesof land acquisition and worsened human rights abuses againstpolitical opponents (Hellum & Derman, 2004; Howard-Hassmann, 2010). Mugabe also blatantly rigged the subse-quent parliamentary and presidential elections, in 2000 and2002 respectively (Bush & Szeftel, 2002; Raftopoulos, 2002).

In response to these events, the EU adopted aid sanctionsagainst Zimbabwe. The EU decision-making process presentsa surprising situation. The UK, which is the former colonialpower with vested interests in the country, lobbied in favorof the application of appropriate and restrictive measures

(Taylor & Williams, 2002, p. 555). On the contrary, France(which had never been a large donor in Zimbabwe) adopteda welcoming approach (Chafer, 2002, p. 352; Cilliers, 2001,p. 124). The International Crisis Group confirms that theEU “has had difficulty speaking with one voice on Zimbab-we”: some Member States (the Scandinavians, the Dutch,and the UK) wanted to act quickly and increase pressure onMugabe; others, like France and Belgium, insisted that timefor action was not yet ripe (International Crisis Group,2002, p. 15).

Variation between France and the UK is also evident if onelooks at their bilateral relationships with the country. Accord-ing to Addison and Laakso, the UK “has been most active inapplying diplomatic pressure” against Mugabe (Addison &

Laakso, 2003, p. 468). Since 2001, DfID had stopped alloperations with the government, and channeled all aidthrough multilateral organizations or NGOs (InternationalDevelopment Committee, 2010, Chapter 4). In contrast,France “extended bilateral development cooperation withZimbabwe at the same time most of the other EU countrieswere drastically reducing cooperation” (Grebe, 2010, p. 13).The soft stance adopted by France is confirmed by otherresearchers: “EU Member States and the EC delegation havecoordinated their positions through Heads of Missions(HOMs) meetings. The UK, for example, organized regularmeetings during its Presidency . . . Such cooperation excludessome Member States: most importantly France, which todayis one of the largest creditors to Zimbabwe” (Laakso,Kivimäki, & Seppänen, 2007, p. 71). In its resolution of

September 6, 2001, the European Parliament took the uncom-mon step to directly criticize the French government, statingthat “whereas most [EU] Member States have substantiallycut or suspended their financial support and developmentaid to Zimbabwe, . . . France had significantly increased itsengagement”, and called on Paris “to adhere to the line takenby the other Member States” (European Parliament, 2001). 8

4. EXPLANATION OF VARIATION

(a) Analytic eclecticism

The complexities of aid decision-making defy simpleexplanations. In one of the most acclaimed overviews of


http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-

8/17/2019 1-s2.0-S0305750X1500011X-main (1)

8/20

contemporary aid, Lancaster (2007, p. 9) denied the possibilityof creating an overarching framework to explain developmentassistance explicitly because “there are too many interactingvariables to justify a model that would be both parsimoniousand insightful”. Chafer and Cumming (2011, p. 9) similarlyargued that French and British bilateral cooperation “cannotbe understood without disaggregating the state[s] and includ-

ing strong focus on the perceptions, interest and ideas of elitepolicy-makers, not to mention institutional constraints andother domestic variables”.

These comments pushed the author to go beyond paradig-matic boundaries and adopt “analytic eclecticism”. Analyticeclecticism acknowledges the complexities and messiness of everyday politics and “explores how diverse mechanisms posit-ed in competing paradigm-bound theories might interact witheach other” (Sil & Katzenstein, 2010, p. 10). This does notimply that “anything goes”: eclectic work needs to be assessedin relation to available evidence and alternative arguments.However, analytic eclecticism avoids the risk that a lack of integration among multiple paradigms can become a “hin-drance to understanding” (Hirschman, 1970, p. 329).

The article concentrates on four factors that can explain a

different attitude to political conditionality across the Chan-nel: material interests, international socialization processes,domestic politics and organizational cultures. 9 Table 3 listsall relevant hypotheses. The section that follows elaborateson their theoretical basis.

(b) Hypotheses

(i) RealismScholars adhering to rationalist and materialist schools of

thought such as realism argue that norms constitute nothingbut ex post rationalizations of strategic interests such as secu-rity and/or wealth (Gilpin, 1983; Waltz, 1979). Following thisline of reasoning, a country’s approach to political condition-

ality would be the consequence of pre-determined priorities interms of (1) favorite recipients and/or (2) sanction decisions. Adonor would appear to internalize political conditionality if the recipient countries that are suggested by its material inter-ests are in the majority good human rights performers, and/orif sanction decisions (which are in reality dictated by its mate-rial interests) can be masked through normative language. Inboth cases, political conditionality would merely representwindow-dressing (for two materialist accounts of aid motiva-tions, see Hook, 1995; Liska, 1960). 10

(ii) International constructivismIn contrast with the realist focus on materiality and ration-

ality, international constructivism argues that the behavior of states is norm-driven (not goal-oriented) and that norms areconstructed (as well as deconstructed) through social interac-tion at the international level (Ruggie, 1998; Wendt,1999). 11 Foreign policies are similar (or dissimilar) when

states construct their identities (and therefore internalize thenorms dictated by these identities) together with (or against)each other (Boekle, Rittberger, & Wagner, 2001). Variationbetween France and the UK can be explained by referenceto the fact that the two donors belong to groups of stateswhich assign different levels of importance to political condi-tionality. Options include:

the Nordic Plus group (which encompasses the UK butnot France), and the Commonwealth of Nations versus the InternationalOrganization of La Francophonie (Organisation Interna-tionale de la Francophonie, OIF).

(iii) LiberalismUnlike realism and international constructivism, and

against long-lasting assumptions dismissing the influence of domestic accountability on foreign policy decisions(Almond, 1950; Lippmann, 1922), liberalism proposes thatstates are “autistic” (that is, they interact badly with thematerial and social international environment). Their foreignpolicies are therefore primarily determined by the interests of dominant local actors (Moravcsik, 1997). 12 Past research hasshown that domestic politics play an important role in aiddecision-making (Jörg Faust, 2008; Milner & Tingley, 2010;Ruttan, 1996), Scholars working on political conditionalityhave specifically suggested four variables that can influencethe application of this norm:

Parliamentary scrutiny. According to Zorbas, aid officialsin Rwanda insisted that their support was not uncondition-

al, for one simple reason: “

We are accountable to our Par-liament and our public” (Zorbas, 2011, p. 109; see alsoWalldorf, 2008, pp. 8–41). Media pressure. Barratt commented that donors “becomeespecially sensitive to their obligation to protect these rightselsewhere when violations are widely publicized” (2008, p.15; van Belle & Hook, 2000, p. 342). Right-wing governments. Faust, Leiderer and Schmittsuggested that the conservative governments which cameto power in Europe in 2009 and 2010 “soon began to assign

Table 3. List of hypotheses to explain variation in the degree of internalization of political conditionality by France and the UK

Research tradition Hypothesis/hypotheses

Realism The UK has internalized human rights to a larger extent than France (1) because the most important recipientsof British aid are better human rights performers than the most important recipients of French aid, and/or (2)because the application of political conditionality could be used by British decision-makers to mask the pursuit of strategic interests in specific countries

International constructivism The UK has internalized human rights to a larger extent than France because it is the object of stronger socialpressure (through persuasion and/or imitation) within (1) the Nordic Plus Group and/or (2) the Commonwealthof Nations (versus the International Organization of La Francophonie)

Liberalism The UK has internalized human rights to a larger extent than France because British aid decision-makers are understricter domestic scrutiny than their French counterparts

Sociological institutionalism The UK has internalized the norm of political conditionality to a larger extent than France because theorganizational culture of DfID presents a more fertile ground than the organizations cultures of the FrenchMinistry of Foreign Affairs and/or the French Development Agency


http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-

8/17/2019 1-s2.0-S0305750X1500011X-main (1)

9/20

more importance to political conditionality and the democ-racy objectives of the instrument” (2012, p. 456; see alsoBrech & Potrafke, 2014, p. 73). Strong human rights NGOs. Spence argued that criticismfrom civil society organizations “matters for policymakersbecause it can strengthen domestic opponents of the foreignaid program and may even jeopardize the political survivalof elected officials” (2014, p. 419).

(iv) Sociological institutionalismSociological institutionalism emphasizes the role of culture

in explaining the behavior of large organizations (DiMaggio& Powell, 1991, pp. 1–41). IR scholars exploited this researchtradition to explore puzzling differences between organizationswith similar functions in similar countries, such as the army(Kier, 1995; Legro, 1997). Past research suggests that thisargument can be adapted to aid organizations, and that thedegree of resonance of international norms with specific orga-nizational cultures can explain the influence of human rightsnorms on development practices. For instance, Sarfatyshowed that organizational culture represents one of the mainobstacles to the integration of human rights in the WorldBank (Sarfaty, 2012). Extensive ethnographic researchbrought to surface a dominant subculture of economists, bothnormatively and practically reluctant to integrate political

and/or legal conceptions into the Bank’s operations. The

general conclusion is that “the ways norms become adoptedand ultimately internalized in an institution largely dependon their fit with the organizational culture” (Sarfaty, 2009,p. 649).

(c) Empirical evidence

(i) RealismThe realist hypothesis is parsimonious. However, it is

unconvincing to explain variation between France and theUK. As shown in Tables 4 and 5, the top 15 and 30 recipientsof British aid generally have lower scores on personal integrityrights as well as political rights and civil liberties than the topFrench recipients. 13

Since it is London, not Paris, that delivers the larger part of its assistance in countries with repressive regimes, there is nomaterial justification for British decision-makers to pay lip ser-vice to political conditionality and pretend consistent applica-tion of aid sanctions.

This simple test is in line with the results of past statisticalstudies. Quantitative analyses have shown that France hasapplied political conditionality to a smaller extent than theUK even if one controls for the confounding effects of eco-nomic, strategic, and post-colonial interests. For instance, Ale-sina and Dollar clarified that their findings “are obtained

holding ‘colonial past’ as constant. Therefore they cannot be

Table 4. Scores of top 30 recipients of French and British aid on political rights and civil liberties (Freedom in the World) during 2000–11

United Kingdom France

Recipient Average score2000–2011


Nigeria 4,17 Morocco 4,63India 2,5 Cote d’Ivoire 5,79

Iraq 6,08 Nigeria 4,17Tanzania 3,5 Cameroon 6,04Bangladesh 3,83 Congo, Rep. 5,13Afghanistan 5,83 DRC 5,96Ethiopia 5,17 Tunisia 5,5Pakistan 5,17 Senegal 2,83Ghana 1,83 China 6,5Uganda 4,71 Iraq 6,08DRC 5,96 Egypt 5,63Sudan 7 Vietnam 6,21China 6,5 Algeria 5,5Malawi 3,67 Madagascar 3,67Zambia 3,79 Indonesia 2,92South Africa 1,75 Lebanon 4,83Kenya 3,75 Turkey 3,33

Mozambique 3,42 Kenya 3,75Rwanda 5,71 Mali 2,33Sierra Leone 3,5 Burkina Faso 4,04Serbia 2,25 Gabon 4,92Nepal 4,25 Niger 3,71Zimbabwe 6,17 Mexico 2,42Vietnam 6,21 Pakistan 5,17Indonesia 2,92 South Africa 1,75Somalia 6,75 Mozambique 3,42Cameroon 6,04 Guinea 5,54Yemen 5,33 Mauritius 1,42Myanmar 6,96 Brazil 2,29Cambodia 5,54 Ghana 1,83Average Top15 4,65 Average Top15 5,10

Average Top30 4,68 Average Top30 4,24

Source: Freedom in the World Comparative and Historical Data (http://www.freedomhouse.org/report-types/freedom-world).


http://-/?-http://-/?-http://www.freedomhouse.org/report-types/freedom-worldhttp://www.freedomhouse.org/report-types/freedom-worldhttp://-/?-http://-/?-

8/17/2019 1-s2.0-S0305750X1500011X-main (1)

10/20

explained by the fact that different colonizers have more or lessdemocratic regimes in former colonies” (Alesina & Dollar,2000, p. 49).

A close look at the case studies presented in the previoussection confirms the weaknesses of the realist hypothesis.Laakso et al commented that “while the UK had a larger con-centration of interests in Zimbabwe than any other EU Mem-ber State, its approach did not stem from those interests butrather from principles and values of the Cotonou Agreement.The British business in the country, for instance, has notbenefited from the smart sanctions” (2007, p. 70; see alsoGallagher, 2011, p. 54; Taylor & Williams, 2002, p. 552). Onthe opposite side, France had limited material interests in Zim-babwe. Yet, this was sufficient to resist the application of sanc-tions. The International Crisis Group reported that France“did not wish to antagonize Mugabe because his supportwas important for the Kabila government in Kinshasa”(International Crisis Group, 2002, p. 15). Some critics evensuggested the relevance of new (but not at all considerable)commercial deals, such as the lease of aircraft to Air Zimbab-we (Shumba, 2003a, 2003b).

Mozambique offers a similar picture. New discoveries incoal and gas have driven the boom of the Mozambican econ-omy (Cuvilas, Jirjis, & Lucas, 2010; Hatton & Fardell, 2012)which in turn has meant that both French and British firms

have shown an increasing interest in the country. For example:

Constructions Mecaniques de Normandie and BNP Paribaswere directly involved in the establishment of the controversialEmpresa Moçambicana de Atum (Korby, Burkhardt, &Pronina, 2013; Roy, 2013); British-Australian Rio Tintobought extensive assets in the country during 2010–12(Marais, 2011; Roberts, 2011); British multinational BritishPetroleum (BP) has pulled out of five countries in SouthernAfrica (namely, Namibia, Malawi, Tanzania, Zambia, andBotswana) but still invests in order to grow its market sharein Mozambique (Reuters, 2010). Notwithstanding similarinterests, France adopted a soft stance during the donor strike,while the UK took the leadership of it. A realist approach is of little help to explain this situation.

(ii) International constructivismThe Nordic Plus Group. Past research has shown that North-

ern European donors (such as, Denmark, the Netherlands,Norway, and Sweden) not only set the highest standards interms of aid sanctions against repressive regimes (Noel &Therien, 1995; Stokke, 1989), but also deliberately trigger pro-cesses of persuasion and imitation at the international level.Nordic countries directly act as “norm entrepreneurs” or“mentor states” (Herman, 2006; Ingebritsen, 2002). In addi-tion, they “consciously see their role in the aid process asone of seeking . . . to reform and correct the aid process by

example”

(Lumsdaine, 1993, p. 66).

Table 5. Scores of top 30 recipients of French and British aid on personal integrity rights (Political Terror Scale) during 2000–11

United Kingdom France



Nigeria 3,88 Morocco 2,71India 3,88 Cote d’Ivoire 3,79

Iraq 4,75 Nigeria 3,88Tanzania 2,63 Cameroon 3,25Bangladesh 3,75 Congo, Rep. 2,87Afghanistan 4,79 DRC 4,83Ethiopia 3,71 Tunisia 2,83Pakistan 4,13 Senegal 2,48Ghana 2,52 China 4,00Uganda 3,79 Iraq 4,75DRC 4,83 Egypt 3,58Sudan 4,96 Vietnam 2,67China 4,00 Algeria 3,71Malawi 2,54 Madagascar 2,50Zambia 2,95 Indonesia 3,54South Africa 3,08 Lebanon 2,92Kenya 3,42 Turkey 3,42

Mozambique 2,96 Kenya 3,42Rwanda 3,13 Mali 1,79Sierra Leone 2,79 Burkina Faso 2,33Serbia 2,00 Gabon 2,42Nepal 4,13 Niger 2,54Zimbabwe 3,75 Mexico 3,33Vietnam 2,67 Pakistan 4,13Indonesia 3,54 South Africa 3,08Somalia 4,33 Mozambique 2,96Cameroon 3,25 Guinea 3,17Yemen 3,73 Mauritius 1,79Myanmar 4,29 Brazil 3,92Cambodia 2,96 Ghana 2,52Average Top15 3,81 Average Top15 3,43

Average Top30 3,57 Average Top30 3,17

Source: Political Terror Scale Ratings (http://www.politicalterrorscale.org/download.php).


http://www.politicalterrorscale.org/download.phphttp://www.politicalterrorscale.org/download.php

8/17/2019 1-s2.0-S0305750X1500011X-main (1)

11/20

In case of human rights or democratic crises, the UK finditself under stronger pressure than France to take similar posi-tions to Nordic donors for two main reasons. First, normativeaffinity. UK representatives feel much closer to their “like-minded” counterparts than French officials do (Olsen, 2011,p. 98). A DfID official highlighted that, “even though theUK and the Nordics might disagree on the appropriate

response to specific abuses, I always trust them to share ourown values and objectives. This is why I pay attention to whatthey say”. 14 Second, closer coordination. The UK and Nordiccountries strictly coordinate on development issues, so muchso that harmonization has been formalized with the creationof the Nordic Plus Group. 15

British examples in support of the international construc-tivist hypothesis can be found at all levels: policy documents,development agreements, and aid sanctions. Policy synchro-nization between the UK and Nordic donors goes back tothe end of the 1990s, when the Ministers for DevelopmentCooperation in the Netherlands, Germany, Norway and theUK were all women, and created the “Utstein Group”(Lawry-White, 2003, p. 4). At that time, the affinity betweenClare Short, the UK Secretary of State for International

Development, and the other Development Ministers meantthat the newborn DfID—eager to position itself as a leaderin the development sector—recurrently looked at the policiesof the Nordic countries to draw inspiration. At the level of international arrangements, the Joint Financing Arrangementpromoted by the Nordic Plus Group explicitly included ahuman rights clause (Nordic Plus, 2007, p. 11). The same istrue (under specific instruction from the Netherlands) for theNordic Plus Group’s template for arrangements on delegatedcooperation (Nordic Plus, 2006, p. 37). In case of abuses,almost all interviewees emphasized a good level of harmoniza-tion with Nordic countries. For instance, in Maputo the maingroup for donor coordination is the G19. However, in reality,the informal coordination among Nordic Plus countries has

more cohesion. An aid official commented: “

the donors whichparticipate in the Nordic Plus Group adopted a similar stanceboth during the donor strike and in response of the EMA-TUM deal. This is not surprising if you think that we considerourselves to be obvious collaborators in numerous circum-stances”. 16

Interviewees and other researchers confirm the absence of any similar “natural partnership” between France and Nordiccountries. Nordic diplomats often mention French unwilling-ness to “coordinate and be coordinated”. A foreign diplomatin Madagascar shared his surprise when the new FrenchAmbassador in Antananarivo, who arrived in the middle of the political crisis, showed no interest in knowing what otherlocal diplomats thought about the crisis and how they wouldhave reacted to the unilateral French decision to continue

development assistance to Rajoelina.17

During the SwedishPresidency of the EU in the second half of 2009, a numberof Francophone countries were on the agenda. The Presidencyfound it very difficult to find common ground among themember states for policy initiatives directed toward these fourcountries and noted that “Swedish decision-makers in generalconsider France as a difficult partner to work with in an Afri-can context” (Olsen, 2012, p. 416).

Lack of persuasion due to poor coordination is accompa-nied by limited chances for imitation. The French diplomatsinterviewed have never demonstrated any deference to theNordic approach. Even when they acknowledged that Nor-dic donors score better than anybody else in most rankingsof aid quality, they differentiated France on the basis of theargument that “we are a much greater power, with larger

interests. In the end, the only interest of Nordic countriesis to be at the top of these rankings. It is not a fair compar-ison”. 18

The International Organization of La Francophonie versusthe Commonwealth of Nations. France and the UK managetheir relationship with former colonies (and a few additionalrecipient countries) through separate international organiza-

tions: the OIF and the Commonwealth. In the case of grosshuman rights abuses and grave violations of democraticprinciples, both the OIF and the Commonwealth can applyaid sanctions (and, more generally, suspend repressiveregimes from membership). In addition, Canada (which isa strong supporter of political conditionality) plays a sig-nificant role in both organizations, thus at least partiallycounterbalancing the hegemonic position of the former colo-nial empires.

Notwithstanding these similarities, two factors makesocialization in favor of political conditionality more likelyfor the UK within the Commonwealth than for France with-in the OIF. To begin with, while both organizations officiallysupport human rights and democratic institutions amongtheir members, there is no doubt that the Commonwealth

has a much stricter policy in this respect (Srinivasan, 1997,p. 513). The different degree of importance given to democ-racy and human rights issues is confirmed by the way inwhich the two organizations reacted when Cameroon appliedfor membership in both of them in 1989: the country’s demo-cratic transition was monitored far more closely and criticallyby the Commonwealth than by the OIF (Torrent, 2011, p.611).

Second, while both organizations are significantly influencedby the former colonial powers, the OIF is heavily France-centered. For instance, Paris is by far the largest contributorto its budget. The Commonwealth is more balanced. Thebudget is evenly distributed, and the organization includesother relevant players (not only Canada, but also Australia

and New Zealand) (Sanders, 2014, p. 25). In addition, “

thereare no ‘permanent members’ in the Commonwealth of Nationsleadership. In the Commonwealth Ministerial Action Group(CMAG), created to be a kind of executive for the Common-wealth, no nation is assured of a place, not even the UnitedKingdom” (Kirby, 2011, p. 56).

The problem with the “OIF versus Commonwealth”hypothesis is that, as a matter of fact, the Commonwealthhas not played a significant influence over British aid deci-sions. In the case of Zimbabwe, for example, the Common-wealth suspended the country from the organization in 2002(Abraham, 2004). However, it was Britain itself that pushedfor the imposition of international sanctions against Mugabe(Taylor, 2005, p. 373; Williams, 2004, p. 916). After the 2006Fiji coup, the Commonwealth did influence the UK position

in favor of aid sanctions. However, several interviewees point-ed to the fact that it was mainly due to the willingness to fol-low Australia rather than the Commonwealth that Britainacted so quickly. 19 At the policy level too, it is often Britainwhich leads the human rights activities of the organization.For instance, it was the UK that lobbied for the expansionof the CMAG remit to address human rights abuses beyondunconstitutional overthrows of governments (Williams, 2005,p. 383), and it was the UK that put the issue of gay rightson the agenda of the organization in 2011 (BBC, 2011).

In sum, the analysis of the case studies and the interviewswith their protagonists do not warrant the conclusion thatFrance and the UK differ in the internalization of politicalconditionality because the f ormer belongs to OIF and thelatter to the Commonwealth. 20


http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-

8/17/2019 1-s2.0-S0305750X1500011X-main (1)

12/20

(iii) LiberalismWhen interviewees were asked to suggest the most impor-

tant factor behind the different attitude to political condition-ality by France and the UK, there was almost unanimousconsensus (from both capitals and local offices) in favor of domestic politics. The case study of Zimbabwe confirms thisinterpretation. According to Vines, it was domestic pressure

that “

forced the [UK] government to take action proactivelyagainst the regime of President Mugabe (Vines, 2011, p. 28.see also Porteous, 2005, p. 291). Taylor and Williams com-mented that “few events in Africa in recent years have so excit-ed British opinion on foreign affairs as the downward spiral of Zimbabwe under President Mugabe” (Taylor & Williams,2002, p. 547).

However, not all the variables suggested by previousresearch (and summarized in the previous sub-section onhypotheses) are supported by sufficient evidence. For instance,the hypothesis that conservative governments should havebeen more in favor of political conditionality than left-wingexecutives has to be rejected because the favorable British atti-tude was actually driven by the New Labour, at the same timewhen the right-wing presidencies of Chirac and Sarkozy

neglected the norm.21

The hypothesis based on the strengthof NGOs is also unconvincing. Most British developmentNGOs have never supported a strict application of politicalconditionality (fearing “double punishment” of the poorpopulation).22

The two crucial domestic variables are media pressure andaid decision-making structure (the latter being something notexplored by previous research). To begin with, national mediadedicates much more attention to aid in the UK than inFrance (Van Belle, Rioux, & Potter, 2004, p. 73). British news-papers (including popular tabloids) frequently cover scandalsof money disbursed to repressive regimes (see the examplesof Groves, 2014; Kelly & Ford, 2012; Oborne, 2010), andrecurrently call for aid cuts at times of austerity or national

disasters (Chapman, 2013; Clark, 2013). The British govern-ment’s attempt to sell its “ethical” foreign policies (both athome and abroad) has also opened the policy process up tonew levels of scrutiny (Gaskarth, 2006, p. 53; Williams,2004, p. 910). The result is that “the potential for a strongdomestic reaction to an overseas development is always thereand policymakers must factor that potential into their think-ing” (Coles, 2000, p. 111).

This is not the case on the other side of the Channel. Eventhough in the middle of the 1990s numerous legal investiga-tions into the affairs of some leading figures of African net-works and France’s role in the Rwandan genocide seemed towake up the French public toward the fate of Sub-SaharanAfrica (Kroslak, 2007; S. Smith & Glaser, 1997), a senior offi-cial at the French Development Agency commented that aid is

still rarely mentioned in newspapers’ headlines and the elec-torate “is largely ignorant about who does and who takeswhat”. 23 Numerous scholars have lamented the lack of trans-parency, and the fact that that the media and public opinionare more often than not “kept out of the game” and “by-passed” (Bovcon, 2013, p. 16; Renou, 2002, p. 10).

With respect the decision-making structure, British aidresponsibilities squarely fall on the shoulders of one singleindividual, the Secretary of State for International Develop-ment, and a few senior officials within one single department,the Department for International Development (Ireton, 2013,p. 50; see also Fisher, 2015). This situation considerablyenhances domestic accountability. For example, former UKSecretary of State for International Development AndrewMitchell waited until his last day in office to restore £16 m

of aid to Rwanda in September 2012 (Hale, 2012). In contrast,France lacks a high-level political figure as Minister of Devel-opment Cooperation. In addition, joint management of theaid budget by the Ministry of Foreign Affairs, the Ministryof Economic Affairs and AFD, and the significant (but oftenhidden role) played by the President make it extremely difficultto assign clear-cut political responsibilities. 24 Naudet (1997, p.

176) commented that French aid lacks transparency andaccountability “partly as a result of the dispersal of the admin-istrative structures involved . . . the interventions are spreadbetween agencies according to their own specific objectivesand there are insufficient overviews available to the publicwhich attempt to place overall activities within a coherentframework”.

In a similar vein, the UK is a parliamentary democracywhere decision-makers are directly accountable to the Parlia-ment. Institutional accountability adds up to political account-ability. DfID decisions are examined by the House of Common’s International Development Committee, a seriousdiscussion over the aid budget takes place every year, andMembers of the Parliament (MPs) often conduct in-depthreview of specific aid programs (see House of Commons,

2013, pp. 219–227). For instance, in 2012 Andrew Mitchellwas “grilled” by MPs over his decision to restore British aidto Rwanda notwithstanding general concerns about thehuman rights record of the country’s government (Mulholl,2012). In contrast with British parliamentary form of govern-ment, France is a semi-presidential system where those incharge of foreign policy and high-level aid decisions respondto the President, not to the Parliament. According to Médard,“the Parliament has never played any important role in Afri-can policy. Moreover, there has never been a real publicdebate in Parliament on French policy. The budget of theMinistry of cooperation was voted on without any serious dis-cussion” (Médard, 2004, p. 42). The absence of formalaccountability channels is exacerbated by the low level of

interest shown by MPs, as confirmed by the absence of a speci-fic committee dedicated to aid issues (development assistanceis discussed within the more general foreign policy committee)(Lancaster, 2007, p. 164).

(iv) Sociological institutionalismNotwithstanding the performance of similar functions, the

institutions managing French and British development assis-tance offer significantly different organizational cultures.French aid is dominated by economists and diplomats. The“pivot” actor of the French aid system is the French Develop-ment Agency (AFD), whose staff is mainly composed of econ-omists who eagerly repeat that “AFD is a bank and operatesas a bank”. 25 Meimon confirms that AFD “is not a politicalorgan” but rather takes a “banking approach” to development

issues (2007, p. 47; Raffinot, 1998, p. 477). The remaining partof French bilateral aid is managed by diplomats within theMinistry of Foreign Affairs (Gabas, 2005, p. 253). In theory,diplomats might be more inclined than economists to assigndue consideration to a political issue like respect for humanrights. Yet, several interviewees suggested that this is not thecase. To begin with, French diplomats value “continuity overdisruption, partnership over imposition”. 26 As clarified by theformer Secretary of State for Cooperation and La Franco-phonie, Alain Joyandet (2010), France speaks “to regimes thatare not perfect democracies. . . . This is a diplomacy of “positive influence”: we want to talk with everyone to bringthem to our values. The more we talk with these countries,the greater the possibilities of advancing these values”. More-over, diplomats are educated under a “Gaullist consensus”,


http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-

8/17/2019 1-s2.0-S0305750X1500011X-main (1)

13/20

which supports the idea that Africa acts as the guarantor of French standing in world politics. Rupture of any significantrelationship with African countries is almost never contem-plated (Bourmaud, 2000).

In principle, DfID offers a more favorable terrain tohuman rights concerns. First, the existence of a dedicatedDepartment in charge of development assistance insulates

aid decisions from political pressures to downplay humanrights in favor of strategic and economic interests (Killick,2005, p. 675; Porteous, 2005, p. 282). Second, DfID is exten-sively staffed with individuals who share a strong ethical pur-pose (as confirmed by the popularity of the Department involuntary application to the Civil Service and the large num-ber of recruits with work experience in NGOs) (Vereker,2002, p. 136). Nevertheless, a cultural explanation of varia-tion between France and the UK presents two serious weak-nesses. First, decisions over aid sanctions are taken atministerial level, not by DfID staff. Second, historicalaccounts of decision-making over aid sanctions indicate thatthe Department which often takes the hardest stance onhuman rights is the FCO, not DfID. For example, at thebeginning of the last decade the FCO “came to feel that

Short was not paying enough attention to human rightsand democratization in countries like Rwanda and Uganda”(Porteous, 2005, p. 286). Rwanda fomented strife also morerecently. When Mitchell reinstated aid to the country, “mostpeople in the Foreign Office [thought] the time [had] come totake a really tough stand” (Groves, 2012).

The result is that, as suggested by several interviewees, a cul-tural perspective is more useful to explain which are the “hardcases” for aid sanctions (that is, those recipient countries towhich a donor find more difficult to apply political condition-ality) rather than the general attitude toward the norm itself.A Nordic official commented: “Look at the countries whereFrance has not applied sanctions: Madagascar, Niger, etc.Then, look at the countries where the UK has been reluctant

to apply sanctions: Ethiopia, Rwanda, etc. In the case of France, strategic interests overcome human rights; in the caseof the UK, development successes overcome human rights”. 27

5. CONCLUSION

Comparative research on political conditionality has exces-sively focused on behavior over discourse, and on assessmentover explanation. This article advances this strand of researchby (1) adopting an original three-fold framework (comprisedof policy documents, international agreements and aid sanc-tions) to measure the degree of internalization of political con-ditionality, and (2) executing a rigorous Most Similar SystemsDesign (or Mill’s Method of Difference) in order to explain

variation between France and the UK.The first part of the article showed that the two donors differ

significantly. While Paris generally opposes political condi-tionality and almost never includes a human rights clause ininternational treaties, London considers respect for humanrights to be a fundamental pre-requisite for long-term partner-ships with developing countries and specifies this condition indevelopment agreements. The analysis of two “hard” cases(Mozambique and Zimbabwe), in combination with pastquantitative and qualitative findings, corroborated that aidbehavior usually reflects the differences highlighted in policydocuments and legal texts. 28

The second part of the article avoided paradigmatic bound-aries and tested four different hypotheses to explain variation

between France and the UK. The UK has internalized politi-cal conditionality to a larger extend than France becausenational media dedicate more attention to aid decisions inLondon than in Paris, and because British aid responsibilitiessquarely fall on one single individual, the Secretary of State forInternational Development, and a few senior officials withinone single department, DfID. These factors ensure the exis-

tence of stricter political accountability. In addition, norma-tive resonance and closer coordination mean that Britishdecision-makers are subject to higher social pressure than theirFrench counterparts when Nordic countries adopt policy posi-tions in favor of political conditionality, insert human rightsclauses in joint financing agreements and apply aid sanctionsin response to human rights abuses and democratic setbacks.Evidence was not sufficiently strong to support the plausibilityof explanations based on material interests or Commonwealthinfluence. Different organizational cultures are not capable of explaining the degree of internalization of political condition-ality, but help detect the recipient countries against which poli-tical conditionality is not applied.

As a synthesis of these arguments, and knowingly runningthe risk of over-simplification, one can conclude that while in

France foreign aid is framed as development cooperation (asthe Minister in charge of development assistance was calleduntil a few years ago), in the UK it is perceived as develop-ment assistance. In the former case, the relationship withdeveloping countries cannot be jeopardized by idiosyncraticproblems (such as human rights abuses) that are bestaddressed through long-term engagement. In the latter case,disbursement of taxpayers’ money must be suspended orwithdrawn to respond significant misbehavior by recipientgovernments.29

In addition, while both countries undeniably use aid alsoas an instrument to foster their national interests (in additionto the flagship goal of reducing poverty), their objectives aredifferent: locally oriented influence for Paris, globally orient-

ed influence for London. While France aims to become atrusted partner to politicians and businessmen in recipientcountries in order to play a significant role in local, regionaland international fora, the UK aims to be perceived as aprincipled actor that follows value-based standards in orderto assume a leadership role in global debates (for a differentbut complementary argument, see Gallagher, 2009). A strictapplication of human rights conditionality obviously endan-gers the goal of the former but can support the ambitionof the latter. 30

Future research should explore additional factors that canexplain different levels of internalization of political condition-ality by bilateral donors. Some interviewees suggested the rele-vance of aid modalities. Legal and economic aspects makeloans more difficult to suspend and/or cancel than grants.

First, a loan is often based on a contract with a third party.As human rights clauses are rarely included in project-levelagreements, aid sanctions would lead to controversial breachesof contractual obligations. Second, discontinuing a loan inevi-tably entails an unforeseen loss for the development agency. Arelated argument proposes that widespread use of general bud-get support (in contrast with project aid) should be accompa-nied by higher levels of political conditionality. First, theinstrument does not entail the presence of workers on theground, and is therefore easy to suspend. Second, the fact thatmoney is disbursed directly into the hands of the recipient gov-ernment calls for stricter fiduciary safeguards (Faust et al.,2012; Hayman, 2011; Molenaers, 2012; Molenaers, Gagiano,Smets, & Dellepiane, 2015).


http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-

8/17/2019 1-s2.0-S0305750X1500011X-main (1)

14/20

NOTES

1. Budget support is aid channeled directly to a country’s budget, to bedisbursed according to its own allocation, procurement, and accountingsystems.

2. Interviewees are anonymous because the commitment not todisclose specific identities was important in order to obtain candidperspectives.

3. Because of space constraints, the assessment of internalization focusesonly on the “negative” aspect of political conditionality, that is, aidsanctions. However, the author acknowledges that “conditionality canmean more than just the enforcement of sanctions . . . Conditionality caninfuse the regular decision-making process over aid allocations, whereby acountry’s inflow is maintained, increased or reduced from previous years”(Burnell, 1994, p. 487). A more detailed theorization of aid conditionalityis offered by Koch (2015).

4. This strategy is in line with aid effectiveness commitments promotingincreased harmonization of development programs. This suggests that, ashighlighted by Del Biondo, international norms on development assis-tance may be contradictory (for instance, political conditionality versuscountry ownership) (Del Biondo, 2015).

5. Comparative studies based on statistical techniques present at leastthree serious flaws (for a complementary argument, see Lancaster, 2007,p. 9). To begin with, they are based on OECD figures that are reportedon an annual basis. This leads to overlook important information. Forinstance, what if a recipient government harshly represses the oppositionin February, development grants are reduced or canceled in April, thesituation improves in September, and disbursements are restored inDecember? OECD figures would inevitably be silent in this case. Second,the dependent variable of all past statistical studies is the total amount of

aid in favor of a specific recipient country. However, when a recipientgovernment violates human rights, donors often decide not to suspend ordiminish aid, but to earmark it for specific human rights programs, orshift disbursement through international or non-governmentalorganizations working within the repressive State. These options affectthe modalities of disbursement, but not the total amount of aid. Lastly,mere menaces to suspend, cut, or cancel foreign aid are occasionallysufficient to obtain policy concessions by recipient governments. In thiscase, aid figures are not affected. However, development financing isconditioned on respect for human rights. More detailed figures from theAidData project can significantly improve statistical studies in thisrespect (Tierney et al., 2011).

6. An anonymous reviewer rightly observed that a comprehensiveexplanation of divergences should take into account the potential

influences related to characteristics of the recipient countries (for anexample of research on political conditionality that looks at the interactionbetween donor and recipient characteristics, see Hackenesch, 2015). This iswhy the two cases of Mozambique and Zimbabwe contrast French andBritish behavior not only one against the other, but also against theactions of other relevant donors in the countries. This strategy is intendedto overcome the problems related to the idiosyncratic feasibility of aidsanctions: it would be unfair to judge France and Britain for rejecting aidsanctions if no other country adopted them; instead, it is interesting todiscover whether, in those cases when sanctions are possible (asdemonstrated by the behavior of other donors), the two countriesapplied them or not.

7. Interviews with foreign diplomats in Mozambique, August–September2014.

8. The article does not examine former French colonies because theyrepresent “easy” cases: it would not be surprising to find that France isoften against the idea of severing its post-colonial aid relationships.

Indeed, anecdotal evidence supports this view. Emmanuel used thebehavior of France in Cameroon as a flagship example of what he called

“donor-patrons”, that is, critical aid providers that refuse to exercise theirleverage and force the recipient to undergo political change (Emmanuel,2010; see also Ebolo, 1998). After the 2009 coup in Madagascar, Francenot only did not fully suspend direct development cooperation andcontinued some aid disbursement (Connolly, 2013, p. 6; Vivier, 2010, pp.126, 162), but also “strongly encouraged the international community totake account of realities on the ground” and “called for the continuationof some development aid to address the declining socio-economicconditions” (D ew ar, Masse y, & Baker, 2013, p. 13; s ee a ls oInternational Crisis Group, 2010, p. 6).

9. Cumming explained variation between France and the UK in the firsthalf of the 1990s through the existence of strong personal and affective ties

between French decision-makers and African leaders. La Françafrique,that is, the informal network of connections between French and Africanpoliticians and businessmen, was contrasted with the strong emphasis thatBritain placed on the formal institutions involved in the formulation of development policy (Cumming, 2001, Chapter 9). This article intentionallyavoids the “emotional” factor emphasized by Cumming because, eventhough it has not completely faded away, it possesses weaker explanatorypower in the new Millennium (Machet, 2012, p. 4). According to Gounin,

“whether the critics of Françafrique like it or not . . . France’s policiestoward Africa today are led by a new generation of officials whose ethos isthat of IMF and EU technocracy, not that of “l’Afrique de Papa”, andwho have had increasing success in normalizing the France-Africarelationship” (Gounin, 2010). In addition, the Françafrique regime hasalready received ample coverage by other scholars and journalists(Bourmaud, 1996; S. Smith & Glaser, 1997).

10. For the sake of clarity, it is important to specify that this article doesnot explore whether material interests affect donors’ (un)willingness toapply aid sanctions against some recipient countries and not others (whichhas already been demonstrated to be the case: Nielsen, 2013). The articletests whether material interests affect the degree of donor internalization of political conditionality in general terms.

11. I call this approach “international” constructivism in order todistinguish it from “unit-level” constructivism, whose focus lies on therelationship between local/domestic norms and the identities, interests andactions of states. The distinction was first proposed by Reus-Smit (2002;for an example, see Katzenstein, 1996).

12. The version of liberalism proposed by Moravcsik is different from thestrand of research which goes under the name of “neo-liberalism”. Neo-liberalism is similar to neo-realism (it is a rational and material theory),but emphasizes the role played by international regimes and institutions infacilitating cooperation in an anarchic environment (Keohane, 1984).Since neo-liberalism is mainly focused on cooperation failures in a goal-oriented world, it is not considered to be relevant to explain variation inthe internalization of political conditionality.

13. This statement may seem in contradiction with the general argumentof the article. It is important to remind that the list of top recipients doesnot take into account the potential ways in which aid is disbursed (forinstance, through NGO channels when the recipient country has arepressive government). The list is used only to give a rough indication of favored recipients.


8/17/2019 1-s2.0-S0305750X1500011X-main (1)

15/20

14. Interview with DfID official, May 19, 2014.

15. At the beginning of the 2000s, the four Nordic donors (Denmark,Sweden, Finland, and Norway) plus Ireland, the Netherlands and theUnited Kingdom formally joined forces under the heading of the “NordicPlus Group”. Since then, the group has gained in importance. In 2006,Selbervik and Nygaard argued that “the Nordic Plus group appears as a

more important point of reference than the Nordic countries per se”

(Selbervik & Nygaard, 2006, p. 51).

16. Interview with aid official in Mozambique, August 22, 2014.

17. Interview with foreign diplomat in Madagascar, September 16, 2014.

18. Interview with French diplomat, March 21, 2014.

19. Interview with foreign diplomat in Fiji, August 28, 2014.

20. As suggested by an anonymous reviewer, social pressure from the USis another potentially relevant variable. Political scientists have alreadyexplained French behavior in francophone Africa as a response to

increased competition from the US (J. F. Clark, 2002; Schraeder, 2000). Inaddition, the US government has taken a principled stance in numerouscases. For instance, the 2001 “Zimbabwe Democracy and EconomicRecovery Act” instructed US officials in international financial institutionsto “oppose and vote against any extension by the respective institution of any loan, credit, or guarantee to the government of Zimbabwe” (Ploch,2011, p. 23). However, significant organizational differences between UKand US aid-managing institutions (Chhotray & Hulme, 2009; Harrigan &Wang, 2011), as well as low levels of formal aid coordination between thetwo countries, justify a stronger focu