Bilateral aid to Tajikistan; donor tactics and responses to troubled governance

ABSTRACT

The amount and form of aid distributed to foreign countries may depend on the contexts and characteristics of the recipient countries. Donors who have aid relations with poorly governed countries, while emphasizing good governance, may have to adjust aid tools to circumvent weak institutional environments. In keeping with the literature on aid in contexts of weak governance, the paper analyses donors’ strategies in Tajikistan – an understudied country with geostrategic importance to traditional donors – showing how bilateral donors provided controlled aid. Donors disbursed aid mainly in the form of projects, a highly controllable aid type. Donors, while clearly interested in Tajik governance, appeared to sidestep politically sensitive areas probably fearing uncalled-for diplomatic frictions with the Tajik leadership. Nonetheless, donors did touch on some politically sensitive issues via non-state channels, enabling them to avoid direct interactions with Tajik authorities. Operating in poorly governed countries can be complicated for bilateral donors: they are accountable to their domestic taxpayers; they have to devise sophisticated aid strategies to achieve aid efficacy, especially in countries also deemed strategically important; and they cannot risk compromising bilateral relations.

KEYWORDS:

• Aid capture
• corruption
• foreign aid
• governance
• Tajikistan

Introduction

Foreign aid is typically justified as a response to recipients’ needs, but its amount, form, and efficacy may vary depending on the contexts and characteristics of the recipient countries. Studies suggest that aid works better in recipient countries with good policy and institutional environments (Burnside and Dollar 2000; Claessens, Cassimon, and Van Campenhout 2009) and that poor governance is a major bottleneck to aid effectiveness (Djankov, Montalvo, and Reynal-Querol 2009; Öhler, Nunnenkamp, and Dreher 2012). In line with this, the mission statements of most donors who stress good governance highlight their application of selective aid based on governance quality (Winters and Martinez 2015). If taken at their word, donors’ decisions around aid disbursement should take the quality of recipients’ governance into account.

Relevant studies indicate mixed findings on donors’ responses to governance quality in recipient countries. Some studies reinforce findings that bilateral donors appear to allocate aid based on the policy and institutional environments of recipient countries and that poor governance negatively predicts overall aid inflows (e.g. Bermeo 2010; Claessens, Cassimon, and Van Campenhout 2009; Okada and Samreth 2012). By contrast, others suggest that aid inflows do not necessarily reflect governance quality and that in some cases corruption levels in recipient countries positively predict aid inflows (e.g. Alesina and Weder 2002; De la Croix and Delavallade 2014; Easterly and Pfutze 2008). An explanation for this rather counterintuitive finding is that those countries which need aid the most tend to have weak governance and lack proper institutions. Thus, donors may be hesitant to withdraw support from those poorly governed countries (Acht, Mahmoud, and Thiele 2015).

Donors concerned about the risk of aid misuse, while intending to keep aid relations with such recipient governments, probably approach the recipients with different aid instruments that enable donors to better respond to weak institutional settings. Studies indicate that donors employ various aid delivery tactics depending on the governance quality of the recipients: Clist, Isopi, and Morrissey (2012) and Nordtveit (2014) show better-governed countries receive more budget support aid; Dietrich (2013) finds donors are more willing to channel aid funds through the recipients’ public institutions in better-governed countries; Winters and Martinez (2015) suggest donors provide more diverse aid to recipients with better governance using a larger number of aid modalities and across many sectors. Those findings indicate donors have the means to respond differently to various levels of governance quality across recipient countries.

Among the recipient countries with doubtful quality of governance, Tajikistan is an interesting context given its curious relationship with donors. The country first drew substantial aid with its 1992–1997 civil war and then with the United States (US) war on terrorism in Afghanistan (Kluczewska 2020). Its geolocation connected with the illegal flows of drugs and other illicit goods contributed to drawing additional donor interests in the country (EU n.d.). While considered to be a corrupt authoritarian state under the ruling presidential family (Hofman and Visser 2021; Kluczewska 2020), Tajikistan has managed to maintain aid inflows averaging round 7% of the gross national income (GNI) over the last three decades (World Bank DataBank n.d.). The country has been understudied on its own in the context of foreign aid (Kluczewska 2020). There are some relevant studies: Abduvaliev and Bustillo (2020) assess overall aid effects on the Tajik economic growth; Kluczewska (2020) examines aid trends over time in the country; and Nakaya (2009) studies Tajik transition to oligarchy and roles of foreign aid in the post-war period. Yet, studies on the detailed donor practices in Tajikistan appear particularly lacking.

With these backgrounds, the paper intends to investigate how bilateral donors delivered aid to a country that is mostly ruled by corrupt political elites yet has geopolitical importance to traditional donors. The paper with its findings aims to draw wider academic attention to the country and Central Asia within the aid literature community. Throughout the paper, the term ‘aid capture’ follows a definition by Dietrich (2013): aid mismanagement by corrupt authorities in the recipient country or aid waste due to a lack of institutional capacity (Dietrich 2013). The term ‘elite group’ follows a definition by Hofman’s study on Tajikistan: one politically affiliated to the ruling regime or with blood ties to political affiliates (Hofman 2016). The rest of the paper proceeds as follows: section two builds a theoretical background; section three introduces the Tajik contexts; section four explains the data source and methods; section five discusses findings from the data analysis; and section six concludes.

Theoretical background

Donors regularly interacting with poorly governed countries may be concerned about the impact of their aid on the recipients’ governance. Some studies linking foreign aid and recipient governance suggest positive associations between the two (e.g. Aronow, Carnegie, and Marinov 2012; Jones and Tarp 2016; Mohamed et al. 2015), whereas others suspect aid is negatively related to the governance and public institutions of recipient countries (e.g. Bräutigam and Knack 2004; De la Croix and Delavallade 2014; Lin, Vu, and Hartley 2020). Particularly with nondemocratic recipient countries, studies indicate aid could hinder governance improvement or even help authoritarian regimes survive (e.g. Ahmed 2012; Cooley and Heathershaw 2017; De Mesquita, Bueno, and Smith 2009). When aid contributes to the nontax revenue of nondemocratic governments, it can enable them to use the revenue to thwart political challenges to their authority or regime. These negative findings imply a potential dilemma for the donors who emphasize governance quality as an important aid objective. Those donors therefore might have to consider adjusting their aid tactics, such as controlling aid fungibility.

Aid fungibility is a phenomenon widely documented in the relevant literature and, depending on the study contexts, has different connotations. Generally, it refers to a degree to which aid allocated to a specific purpose is reallocated to finance a different purpose (Rana 2021). For instance, at a sector level, fungible aid can be redirected to other sectors when the aid is earmarked for a specific sector. At a government level, fungible aid can be substituted for the recipient’s public spending when the donor’s intention is to supplement it (Kaya and Kaya 2020; Rana and Koch 2020). Aid fungibility is often suspected to reduce aid effectiveness (Rana and Koch 2020). This negative connotation might stem from the donor’s perspective that aid diverted from the donor’s intention would likely be used less productively (Pettersson 2007). Yet some studies on aid fungibility suggest otherwise. For example, Rana and Koch (2020) find fungibility can have positive effects with valid aid reallocation in cases that a marginal value added to the alternative sector is higher than that of the donor’s intended sector (Rana and Koch 2020). Pettersson (2007) finds little evidence supporting that non-fungible aid leads to better economic outcomes (Pettersson 2007).

Nonetheless, donors operating in poorly governed countries are often concerned about aid reallocated by the recipient governments (Bermeo 2016). Under poor institutional environments, more fungible or discretionary aid allows corrupt authorities to access aid funds more easily and redirect the funds for their benefits. This could unintentionally lead fungible aid to reduce the likelihood of improving governance in the corrupt government (Masaki et al. 2021; Wright 2009). Therefore, if the potential negative impact of fungible aid is perceived to be high, donors may prefer to provide less fungible aid to the recipient, the degree to which likely depends on the donor’s strategic relationship with the recipient government (Bermeo 2016). Yet Morrissey (2006) argues this fungibility concern of donors becomes less relevant when aid allocation preferences between the donor and the recipient are aligned, i.e. the recipient government allocates aid more or less the same way as the donor desires (Morrissey 2006).

Of the highly fungible forms of aid, general budget support has been advocated since the late 1990s with growing emphasis on recipient ownership and accountability (Dornan 2017). General budget support refers to financial assistance to a recipient government’s budget, using the recipient’s own financial management system and budget procedure. It provides the recipient government with little restrictive funding (Molenaers, Gagiano, and Smets 2017), even if it comes with a set of policy conditions (Winters and Martinez 2015). Accordingly, general budget support is often perceived as donor trust in the incumbent government of the recipient country and as donor endorsement of the incumbent’s policy choices. Its symbolic leverage thus far exceeds that of any other aid types (Molenaers, Gagiano, and Smets 2017). To donors, general budget support can be highly costly for several reasons: the recipient government may pursue aims that are at odds with the donor’s goals (Masaki et al. 2021); it is harder for the donor to monitor budget-support spending once the fund enters into the recipient’s system; and the fungible nature of budget support makes it highly prone to aid capture (Winters and Martinez 2015). Given its potential risks on donors’ side, budget support is used quite selectively for recipients with valid records of governance quality or for those who have taken required prior actions. In some cases, donors may suspend general budget support as a sanctioning device to send a strong signal to the incumbent government to address issues at stake (Molenaers et al. 2015). Budget support suspension may also indicate a donor’s request to renegotiate aid terms with the recipient government. Donors would not use the suspension lightly unless serious governance issues are raised in the recipient country, such as large-scale corruption scandals, major human rights violations, or wide-spread electoral frauds (Molenaers, Gagiano, and Smets 2017). The United Kingdom (UK), for example, ceased budget support in Uganda due to the 2012 massive corruption scandal involving the Office of Prime Minister of Uganda (DFID n.d.).

Masaki et al. (2021) assert donors likely strengthen their political clout on the recipient’s policies when they entrust aid funds to the recipient’s public system as with general budget support (Masaki et al. 2021). Steinwand (2015) similarly suggests there is a strong private goods component for donors because aid going straight into the recipient’s public system affords donors more direct influence on the recipient’s political decision-making (Steinwand 2015). In this sense, donors likely have a baseline preference for state-to-state aid transfer (Dietrich 2013). However, when the institutional quality is estimated to be low and the risk of aid capture seems high, donors may perceive state-to-state aid transfer implausible. The donors then should make a choice between the recipient’s state channel and non-state channel to deliver aid. But donors may or may not have aid channel choices. If the recipient government does not have a reliable public financing system, the only channel option for the donor would be the non-state channel.

Dietrich (2013) finds the donor channel decision appears deliberate with the quality of the recipient’s governance and institutions. Donors bypass recipient governments when their quality of governance and institution is low because poor governance probably alerts donors to the possibility of aid capture. Therefore the donors may seek out alternative partners allowing them to restrict aid misused by the recipient (Dietrich 2013). Similarly, Acht, Mahmoud, and Thiele (2015) provide evidence that bypassing recipient governments is a donor’s response to the weak state institutions of recipients (Acht, Mahmoud, and Thiele 2015). The study further shows bypassing is more targeted towards aid sectors where the degree of substitutability between aid delivery channels is high. In other words, bypassing the state channel is more likely prevalent in aid sectors not requiring much coordination with the recipient government. Bypassing however is less likely prevalent in sectors requiring sustained strong engagement with the recipient government, such as donor-assisted large-scale infrastructure construction (Acht, Mahmoud, and Thiele 2015).

From a different perspective, Allen and Flynn (2018) explore how the domestic politics of donors influence their aid delivery channel decisions – assuming both liberal and conservative governments see political values in aid and simply use it differently (Allen and Flynn 2018). The study argues left/liberal donor governments are more inclined to channel aid funds through non-governmental organizations (NGOs) and are more likely to prioritize direct poverty alleviation than supporting the recipient government’s agenda. In contrast, right/conservative donor governments place a greater emphasis on channeling aid funds through the recipient government to promote the geopolitical and economic interests of the donor’s constituents (Allen and Flynn 2018). Dietrich (2016) also argues that to understand different aid policies of donors, their political economies need to be understood (Dietrich 2016). The study suggests donors emphasizing market efficiency may outsource aid delivery in poorly governed recipient countries to improve the likelihood of aid reaching the intended beneficiaries. By contrast, donors whose political economies emphasize a strong state in service provision may support aid disbursement through the state. Dietrich (2021) discusses when neoliberal donors distribute aid, they favor non-state partners since the donors view the role of aid as direct efforts to enhance the lives of the poor. On the other hand, when statist donors distribute aid, they favor public-sector partners since the donors see the role of aid as catalytic for overall growth (Dietrich 2021).

Taking the literature into account, donor governments that interact with weakly governed recipient countries and relate governance issues to the higher likelihood of aid capture may have to consider different aid tools while factoring in their own characteristics and preferences. To proceed with the Tajik case, the following section explains the contexts of Tajikistan in foreign aid and governance.

Tajik contexts in aid and governance

Upon the collapse of the Soviet Union, Central Asia became a significant recipient of development aid. Tajikistan has received foreign aid, in particularly high ratios to its GNI, around the late 1990s and early 2000s (Table 1). Two incidences drew considerable donor attention to the country – the 1992–1997 civil war and the US war on terror in Afghanistan (Kluczewska 2020).

Table 1. Net official development assistance to Tajikistan with its ratio to gross national income.

Year	1992	1993	1994	1995	1996	1997	1998	1999	2000	2001	2002	2003	2004	2005	2006
Total Amount^‡ (USD million)	19.4	41.2	98.3	87.5	138.4	123.1	228.2	160.6	166.1	248.8	215.7	177.2	276.4	265.1	234.0
ODA/GNI %^#	0.6	1.6	5.1	5.5	10.7	9.6	12.5	10.7	13.6	15.8	13.1	9.0	11.3	8.5	6.0
Year	2007	2008	2009	2010	2011	2012	2013	2014	2015	2016	2017	2018	2019	2020	2021
Total Amount (USD million)	204.3	235.4	299.3	387.4	262.7	376.6	372.6	339.2	460.4	385.0	343.3	407.9	370.2	768.0	519.6
ODA/GNI %	4.2	3.5	5.0	5.6	3.4	4.0	3.5	3.2	4.4	4.5	3.8	4.5	3.8	8.1	5.1

Source: World Bank Data, ^‡Total amount including both bilateral and multilateral (constant price 2021 USD million), ODA: Official Development Assistance, GNI: Gross National Income.

Literature describes the Tajik civil war as a power struggle in the complex combination of regional backgrounds, sub-ethnic groups, elite networks, and loyalty to the legacy of the Soviet regime. It left devastating impacts on the country, claiming at least 50,000 lives with a tenth of its population internally displaced (Kevlihan 2016; Nourzhanov 2005). During the civil war, donors were mostly involved in humanitarian aid. At the same time, the United Nations (UN) was brought in to mediate a peace agreement as part of international efforts to mitigate the conflict. Following the UN-mediated peace agreement in 1997, donors started engaging in post-war peace and state building with many opening offices in the country (Kluczewska 2020).

Between the late 1990s and early 2000s, prominent donors including the International Monetary Fund (IMF), World Bank, and Asian Development Bank mainly focused on macro-economic liberalization, creation of legal environments, and democratization of the country. Critics, however, argue the donors’ push for institutional reform held back the implementation of the 1997 peace agreement and consequently contributed to maintaining the authoritarian regime through political power consolidation (Kluczewska 2020; Nakaya 2009).

Foreign aid to Tajikistan experienced a shift with the US war on terror in Afghanistan in the aftermath of September 11, 2001. Sharing its entire southern border with Afghanistan, Tajikistan became an essential strategic ally to the US and European Union (EU) by providing an alternative route for the International Security Assistance Force, a military mission in Afghanistan (Boonstra and Shapovalova 2012; Kluczewska 2019a; Kluczewska 2020). During this time, donors also paid more attention to Tajik governance due to fears that a failure of the state could raise a level of insecurity in Central Asia and beyond (Boonstra and Shapovalova 2012; Kluczewska 2020).

Over the last three decades, Tajikistan has been under the leadership of Emomali Rahmon. First elected in 1994, he won five consecutive presidential elections including the 2020 election – none of which were deemed democratic by western monitors (RFE/RL's Tajik Service 2020a). Under his ruling, the constitution was amended several times allowing the president to run for office unlimited times. In 2015, the parliament granted Rahmon the title ‘Leader of the Nation’ making him essentially immune to prosecution and enabling him to secure his dominance (RFE/RL's Tajik Service 2020a; 2020b). There has been speculation that his eldest son, Rustam Emomali, will succeed him as the constitution was amended again in 2016 to lower the age threshold to 30 years for presidential candidates (RFE/RL's Tajik Service 2020a).

The country has been criticized for its massive crackdown on government critics and dissidents, heavy censorship of the media, repression of civic activities, and rampant corruption – all indicative of poor governance (HRW Tajikistan n.d.; World Bank Databank n.d.). Particularly with corruption, the country was ranked in the bottom nine percentile from 2002 to 2022; higher numbers correspond to higher ranks (World Bank Databank n.d.). One notable example to show the country’s corruption status is the court case involving the Tajik Aluminum Company (TALCO) and the presidential family. The case concerning money-laundering and off-shore companies in the British Virgin Islands was settled in 2008 without revealing details, forcing Tajikistan to spend USD 150–160 million in legal fees. This was one of the most expensive court proceedings at the London High Court, equivalent to 5% of the country’s GDP at that time (Nakaya 2009). Even though its true ownership is the subject of intense debate, TALCO has long been suspected of exclusively benefitting the presidential family and its associates (Hryniuk 2018). Some major donors such as IMF and US responded to the case by calling for an independent audit and major reform of the company (Helmer 2008), but foreign aid continued flowing in to the country (Table 1). Against these contexts, the following sections examine how bilateral donor governments provided aid to Tajikistan.

Materials and methods

Using the aid data, the paper analyses the bilateral aid profile of the Development Assistance Committee (DAC) bilateral donors who report aid records annually to the Creditor Reporting System (CRS). The analysis did not include non-DAC donors who are not obligated to report aid records. The aid disbursement records on Tajikistan between 2010 and 2021 were extracted from the Organisation for Economic Co-operation and Development (OECD) statistics (OECD Statistics n.d.). The study period was set based on the data availability matching the study objective. Data in the earlier years were incomplete or missing at the aid type and channel level. The basic unit of the analysis was each annual aid disbursement, defined in the previous study (Acharya, De Lima, and Moore 2006).

For the analysis of aid sector, aid type, and aid channel, the annual aid disbursement records were sorted by the CRS codes. The aid sector analysis was conducted by sorting the aid records with the CRS numeric purpose codes. The aid type analysis was conducted with the CRS capital-alphabet aid type codes. The aid delivery channel analysis was conducted with the CRS numeric channel identification codes. Results of the aid profile analysis focused mostly on the significant donor group, defined by Gehring et al. (2017) as a group of donors cumulatively accounting for around 90% of the total aid inflow in a recipient country (Gehring et al. 2017). Before proceeding, it should be noted some numeric results are not shown in the tables for brevity but are explained in the relevant contexts.

Results and discussions

Bilateral aid profile and significant donors in Tajikistan

During the last decade, a total of 28 DAC donors disbursed aid (Table 2) exclusively as grants. Standard loans consisted of less than 3% of the total disbursement, most of which France and Germany provided. Five donors – Germany, Japan, Switzerland, UK, and US – comprised the significant donor group, contributing to 92% of the total disbursement (Table 2). Of the significant donors, US was the largest contributor, accounting for 28% of the total amount. US began assisting Tajikistan in 1992 soon after its independence from the Soviet Union, and it has since remained as the largest aid provider to the country (USAID n.d.). US may fit Steinwand’s definition of a lead donor: a lead donor has a long stable exclusive relationship with a recipient country in which the donor continues acting as the largest aid provider with no or few donors vying for the role of the top aid contributor (Steinwand 2015).

Table 2. Annual aid disbursement to Tajikistan, its significant donors and top aid sectors.

Year	Yearly aid value^‡	Significant donor※	Value#		% total value#
2010	171.1	US	497.7		28.2
2011	153.2	Germany	376.4		21.4
2012	165.8	Japan	327.8		18.6
2013	154.5	Switzerland	284.9		16.2
2014	184.7	UK	129.1		7.3
2015	153.1	Significant donor total	1615.9		91.7
2016	129.4	Top sector§	Value	% total value	No. of donors
2017	124.5	15	315.0	17.9	26 out of 28
2018	150.8	12	233.3	13.2	18
2019	119.8	11	157.4	8.9	23
2020	120.8	31	154.8	8.8	16
2021	134.4	21	151.7	8.6	5
Total value	1762.2	Top five sector total	1012.1	57.4	–

Note: ^‡Constant prices (2021 USD million), ※ 28 donors: Australia, Austria, Belgium, Canada, Czech Republic, Denmark, Finland, France, Germany, Greece, Hungary, Ireland, Italy, Japan, Korea, Lithuania, Luxembourg, Netherlands, Norway, Poland, Portugal, Slovak Republic, Slovenia, Spain, Sweden, Switzerland, United Kingdom (UK), United States (US), # Accumulated value from 2010 to 2021, and total value USD 1762.2 million, §11 (education), 12 (health), 15 (government & civil society), 21 (transport & storage), 31 (agriculture).

Overall, the donors seemed to be gradually reducing their assistance to Tajikistan (Table 2). Among the donors, UK reduced its aid noticeably. It disbursed USD 16.4 million on average per year between 2010 and 2015, while the yearly average dropped to USD 5.1 million between 2016 and 2021. Although relatively less drastic than UK, Germany reduced its aid from USD 36 million on average per year across the earlier period to USD 26.6 million across the later period. Germany stated it is phasing out its bilateral development cooperation with Tajikistan to be aligned with the Germany’s 2030 aid reform strategy, which does not include the country as a prioritized bilateral partner (Germany Federal Foreign Office n.d.). In comparison, UK, despite its drastic reduction in aid, keeps Tajikistan in its recipient pool for bilateral programs (UKAID n.d.). Of the significant donors, Japan was the only donor that increased its aid to Tajikistan. A document published by Japan noted Tajikistan is strategically important in part due to the Tajik water sector as well as the regional stability of Eurasia, indicating no signs of aid reduction in the coming years (MOFA 2023).

For sectoral aid, the top-supported sectors (Table 2) were mostly determined by the specific sectoral interests of the significant donors (Data not shown). The most supported sector was government & civil society (hereafter, governance). US contributed to 49% of the total amount this sector received, with 25 other donors supporting it as well. While this showed the collective donor interest in Tajik governance, it is possible the significant donors’ sectoral interest influenced the small donors’ practices. Small donors not necessarily seeking aid visibility might refer to the large donors’ sectoral choice since the sector prioritized by large donors likely reduces uncertainties, which may decrease operational costs for small donors.

To Tajik elite groups, the bilateral donors’ intense interest in good governance could be unwelcome because the western concept of good governance includes promoting democracy, accountability, transparency, rule of law, and strong civil participation (Kluczewska 2019b), most of which can compromise their political privilege. Indeed, bilateral donors appeared wary of their potential political intrusion since they tended to avoid highly sensitive governance areas. The donors invested the largest share of governance aid funds (34% of the total aid to governance) in legal and judicial development (sub-sector code, 15130), followed by removal of land mines and explosive remnants of war (10% and sub-sector code 15250). In legal and judicial development, a program financed by the largest donor (US) supported international narcotics control and law enforcement projects during the study period (USDS n.d.).

Dietrich (2011) argues donors distributing aid in highly corrupt countries look for specific sectors to use aid more effectively, whereas corrupt recipients favor sectors that cost less for cooperation compliance, helping them secure additional aid inflows on their minimum achievement of donor demands (Dietrich 2011). With this view applied to Tajikistan, satisfying aid terms in legal development and landmine removal might cost the Tajik elite groups lower in political terms than in politically sensitive areas such as anti-corruption campaigns and election fraud prevention. On the donors’ part, they might prefer not to directly address sensitive governance sectors due to uncalled-for diplomatic frictions. The cautious stance of the donors may be more conspicuous if their needs to strengthen political leverage with Tajik authorities are high for strategic reasons.

For aid types, the following findings suggested the donors gave the Tajik government limited leeway to access aid funds: the country received virtually no budge support (A type); most of the pooled funds (B type) came earmarked; the majority of the total disbursement was provided as projects (C type) (Table 3). Notably with the project-type aid, its proportion was far higher in Tajikistan (85%) than that of the DAC donors’ average across all their recipients (53%) during the study period (OECD Statistics n.d.).

Table 3. Aid disbursement of each aid type and top aid sectors in project type aid.

Aid type (Code)	Value^‡		% of total	Top sector in project type	Value	% project type total
Budget support (A)	0.02		0.0	15	262.1	17.6
Core contributions and pooled programs and funds (B)	135.5		7.7	12	210.0	14.1
	Core	35.5	26.2
	Earmarked	100.1	73.8	21	151.4	10.1
Project-type interventions (C)	1492.2		84.7	31	142.7	9.6
Experts and other technical assistance (D)	56.2		3.2	14	128.0	8.6

Note: ^‡Constant prices (2021 USD million), §12 (health), 14 (water supply & sanitation), 15 (government & civil society), 21 (transport & storage), 31 (agriculture).

Regarding project-type aid, Dreher, Minasyan, and Nunnenkamp (2015) argue donors prefer projects when they have low confidence in the recipient government using the aid effectively (Dreher, Minasyan, and Nunnenkamp 2015). Winters and Martinez (2015) similarly show bilateral donors give proportionally more project-type aid to poorly governed recipients (Winters and Martinez 2015). Related to those findings, the much higher portion of project-type aid in Tajikistan could imply the level of donors’ confidence in its institutional quality. Furthermore, studies suggest sectoral choices for projects depend on the governance quality of recipients (e.g. Acht, Mahmoud, and Thiele 2015; Dietrich 2011; Winters and Martinez 2015). For instance, donors tend to consider large infrastructure projects particularly susceptible to corruption and embezzlement through procurement fraud (Acht, Mahmoud, and Thiele 2015; Dietrich 2011; Winters and Martinez 2015). Based on this finding, details of project sectors may provide additional information about the donors’ view on the Tajik governance quality.

The donors generally favored non-infrastructure sectors, except the transport & storage sector that regularly involves infrastructure components (Tables 2 and 3). Almost all disbursement to this sector (97%) was invested in road construction/maintenance and air transport. The transport sector however was supported exclusively by Japan who stated one of its priority areas in Tajikistan is transport (JICA n.d.). Overall, a more fine-tuned investigation is required to reasonably conclude that the donor community (excluding Japan) avoided infrastructure projects intentionally due to the weak institutional settings. For instance, the nature of grant aid may have limited the donors to finance large-scale infrastructure projects because such projects could be financed by concessional loans with their terms agreed.

Aid delivery channel choices

The donors utilized various aid delivery channels to disburse aid funds (Table 4), and the degree of doing so appeared to depend on the amount of aid funds. On average, the significant donors employed approximately six channels out of seven, whereas the ten smallest donors used approximately two channels. Acht, Mahmoud, and Thiele (2015) suggest smaller donors probably utilize fewer delivery channels to minimize transaction costs compared to larger donors (Acht, Mahmoud, and Thiele 2015). In fact when the donor’s total disbursement was USD 0.2 million (Ireland in Tajikistan), there might have been no practical reason for using multiple channels.

Table 4. Aid disbursement via different channels, and top aid sectors of state channel and non-state channel.

Channel code#		Value^‡		% of total aid		State channel sector	Value	% state total	Non-state sector	Value	% non-state total
State	10000	839.2	839.2	47.6	47.6	21	145.1	17.3	15	223.3	24.2
Non-state	20000	922.9	347.4	52.4	19.7	12	96.5	11.5	12	136.9	14.8
	30000		4.7		0.3	15	91.6	10.9	31	111.8	12.1
	40000		265.7		15.1	23	83.2	9.9	14	96.7	10.5
	50000		59.9		3.4	11	78.1	9.3	11	79.3	8.6
	60000		127.8		7.3
	90000		117.4		6.7

Note: ^#10000: Public sector institutions, 20000: non-governmental organizations (NGOs) and civil society, 30000: Public-Private Partnerships (PPPs) and networks, 40000: Multilateral organizations, 50000: University, college or other teaching institution, research institute or think tank, 60000: Private sector institution, 90000: Other, ^‡Constant prices (2021 USD million), §11(education), 12 (health), 14 (water supply & sanitation), 15 (government & civil society), 21 (transport & storage), 23 (energy), 31 (agriculture).

For discussion clarity, the aid delivery channels were divided into the state channel (code 10000, public sector institutions) and the non-state channel (the rest of the channels). The bilateral donors disbursed 48% of total aid funds via the Tajik state channel and 52% via the non-state channel (Table 4). Through the state channel, the infrastructure sectors (code 21, transport & storage, and code 23, energy) received large portions of the total state channel disbursement (Table 4). Within the transport & storage sector, the largest funding went to road transport (sub-sector code 21020), and within the energy sector, the largest funding to hydro-electric power plants (23220). To implement those infrastructure components, strong engagement of the recipient government may be inevitable regardless of its governance quality.

The governance sector was supported via both channels (Table 4). Through the state channel, legal and judicial development received over 90% of governance funds. Through the non-state channel, aid was dispersed across four sub-sectors: democratic participation and civil society (sub-sector code 15150), decentralization and support to subnational government (15112), human rights (15160), and legal and judicial development. The former three sub-sectors (democratic participation, decentralization, and human rights) might be some of the sectors in which Tajik authorities would be far less comfortable to be involved. The bilateral donors thus probably chose the non-state channel to circumvent direct interactions with the Tajik government to support those sub-sectors. Relatedly, US indicated it would address more politically sensitive issues such as democracy and governance transparency in the context of Tajik civil society and media (USDS 2022). This could be the donor’s indirect approach to the governance issues, while minimizing the possibility of provoking Tajik authorities. In short, the findings appeared to be congruent with the previous studies (e.g. Acht, Mahmoud, and Thiele 2015); bilateral donors utilized the state channel for the sectors/components that require stronger engagement with the recipient government, whereas they bypassed it for the sectors/components that have higher channel substitutability or include politically sensitive issues.

While the collective channel choice represented that of the bilateral donor community, each significant donor showed a distinct channel preference (Table 5). Germany and Japan preferred the state channel. Switzerland, UK, and US favored the non-state channel. The channel preferences of these donors were also closely aligned with their national-level aid channel preferences. Across all its recipient countries, Germany disbursed 68% of its aid via the state channel, and Japan did 85%. Switzerland, UK, and US did 39%, 29%, and 38% respectively, through the state channel (Lee 2022). Dietrich (2016) argues donors emphasizing market efficiency (e.g. US and UK) prefer to contract out aid delivery in poorly governed countries, whereas donors emphasizing a strong state in service provision (e.g. Germany and Japan) support state provision (Dietrich 2016). Taking the above findings together, the significant donors seemed to make their aid channel choices based on their channel preferences as well as the characteristics of aid sectors and Tajik contexts. This then points to that the donors’ channel choices may not necessarily reflect the recipient’s governance quality per se, but rather the choices can be partly determined by who the significant donors are and what sectors they prefer to invest in or avoid.

Table 5. Aid disbursement channel choice of each significant donor.

Channel #	Germany		Japan		Swiss		UK		US
	Value ‡	% of total	Value	% of total	Value	% of total	Value	% of total	Value	% of total
State total	332.9	88.4	259.7	79.2	30.9	10.8	49.0	38.0	102.9	20.7
Non-state total	43.4	11.6	68.1	20.7	254.1	89.1	79.9	61.9	394.9	79.3

Note: ^‡Constant prices (2021 USD million), # 10000: Public sector institutions, 20000: non-governmental organizations (NGOs) and civil society, 30000: Public-Private Partnerships (PPPs) and networks, 40000: Multilateral organizations, 50000: University, college or other teaching institution, research institute or think tank, 60000: Private sector institution, 90000: Other.

Of the non-state channels, the NGO channel is worth examining: the channel received the largest amount (Table 4); and NGOs are directly affected by the Tajik governance (HRW n.d.). For bilateral donors, NGOs are an important alternative or complementary aid channel because they provide the donors with specialized knowledge on local issues, grassroots networks, and a secure foothold to contact the people the donors intend to help (Allen and Flynn 2018; Dietrich 2013). Yet NGOs operating in harsh institutional settings can be scrutinized for political reasons with their activities being severely restricted.

Depending on local circumstances and aid tactics, donors may partner with NGOs based in their country, NGOs based in the recipient country, or international NGOs. In Tajikistan, the bilateral donors partnered mainly with NGOs based in their country, while rarely collaborating with local NGOs (Table 6). The low level of donor partnership with NGOs based in recipient countries is not isolated to Tajikistan; DAC donors granted about ten times more funds to NGOs based in their countries than to NGOs based in recipient countries (DAC 2022).

Table 6. Aid disbursement via each NGO type.

NGO Channel (code 20000)	Value^‡	% of NGO channel total
Donor country-based NGO	253.6	73.0
International NGO	55.9	16.1
Developing country-based NGO	24.6	7.1

Note: ^‡Constant prices (2021 USD million), NGO: non-governmental organizations and civil society.

Arguably, Tajik NGOs are operating under difficult circumstances. They are subjected to intense political scrutiny, have negligible influence on policy decision-making, and rely on donor funding in the near absence of alternative financing sources (USAID 2015; Kluczewska 2019a). Related to Tajik NGOs’ dependence on donors, critics argue some donors’ practices make Tajik NGOs’ activities more complicated because the donors delegate politically sensitive activities to achieve their agendas, such as bottom-up democratization and human rights protection. To implement those activities on the donors’ behalf, Tajik NGOs are to challenge the authoritarian regime under strict censorship and social norms that connect disagreeing with authorities to violence (Kluczewska 2019a).

During the study period, the largest funding for local NGOs went to removal of land mines and explosive remnants of war. This finding should not serve as an indication that there was little evidence for the donors’ delegating sensitive issues to Tajik NGOs since this paper did not examine individual projects with their sub-contractors. The negligible partnership between the donors and Tajik NGOs might be the outcome of several factors: sociopolitical environments that limit the scope of NGO activities; the donors’ desire to avoid unnecessary conflicts with Tajik authorities by partnering local NGOs; and their competency compared to NGOs based in donor countries or international organizations.

Conclusions

Tajikistan, nestled in Central Asia and ruled by an oppressive authoritarian regime, has continued to receive aid over the last three decades due initially to its humanitarian crisis and later due to its geopolitical importance to traditional donors. Yet research focusing on detailed donor practices in the country has been lacking. Using the aid profile records, the paper explored how and where the bilateral donors disbursed aid in Tajikistan. The overall findings indicate the donors appeared to control fungibility of their aid funds, mainly by disbursing the funds in the form of projects (aid type) in non-infrastructure sectors (aid sector). For the aid delivery channel, the choice made by the donors might not necessarily reflect solely the Tajik governance quality since each donor’s own preference as to aid channel and sector seemed to influence its channel choice.

The donor governments, while strongly interested in Tajik governance, took a cautious stance when providing governance-sector aid through the state channel, probably to avoid potential diplomatic frictions. For politically sensitive governance agendas, the donors did touch on them via the non-state channel. This possibly allowed the donors to sidestep direct interactions with the Tajik leadership. In this sense, the donors’ low partnership with local NGOs could be the extension of their cautious stance, given that the Tajik NGO community is subject to strict political censorship. However, it is uncertain whether this donor cautiousness would be observed in other weakly governed recipient countries that are of little strategic importance especially to traditional donors.

For bilateral donors, operating in poorly governed countries can be complicated because the donors are accountable to their domestic taxpayers for the public spending on such recipient countries. In addition, bilateral donors interacting with recipient countries that are deemed strategically important may have to devise more sophisticated aid strategies that are able to achieve aid efficacy whilst not compromising bilateral relations. Therefore, those bilateral donors may need to allocate more resources to formulate better approaches for aid.

Regarding strategic importance of a recipient country, it would be interesting to investigate whether and how bilateral donors change their aid practices under the evolving geopolitical situations surrounding Tajikistan and the region: China’s aggressive economic investment in Central Asia, US’ withdrawal from Afghanistan in 2021, and Russia’s less predictable regional clout. This would lead to a better understanding about the influence of geopolitical importance in donor-recipient dynamics. Also of interest would be to examine whether there have been underlying interactions among donors themselves. These interactions might open (e.g. competition) or limit (e.g. cooperation) an opportunity for aid capture, leading to a better understanding about the influence of donor-donor dynamics.

Disclosure statement

No potential conflict of interest was reported by the author(s).