Foreign aid during the COVID-19 pandemic: evidence from Turkey

ABSTRACT

Turkey provided medical aid to more than 70 countries during the first months of the COVID-19 pandemic. This article explains the distribution of Turkish medical aid across countries relying on a novel dataset which draws on online news published by Turkey’s state-run news agency and a pro-government news outlet between February 1^st and July 31^st, 2020. The findings show that receiving Turkish medical aid is extensively related to the determinants underlying Turkey’s foreign aid policy. In particular, recipient countries’ historical ties to the Ottoman Empire, their level of imports from Turkey, and their needs based on the state of their general health system are seen to be significant predictors of receiving Turkish medical supplies during the coronavirus pandemic. Regarding partial evidence of cultural similarity, Turkic identity is a predictor of receiving medical aid, whereas Islamic affinity does not have a significant association. Therefore, I find that Turkey has adopted a selective strategy in response to an indiscriminately damaging crisis while keeping a majority of the elements of its foreign aid repertoire in play.

KEYWORDS:

• COVID-19 pandemic
• coronavirus
• foreign aid
• foreign policy
• medical aid
• Turkey
• Turkish foreign aid

Introduction

‘There is hope after despair and many suns after darkness.’ This uplifting quote by the 13^th-century Sufi mystic Rumi appeared on the majority of Turkey-prepared medical aid packages sent to a total of 72 countries during the first months of the COVID-19 pandemic. Yet it is puzzling why some countries received medical aid from Turkey and some did not. Drawing on a dataset recording Turkey’s global medical aid (TurAid) during the COVID-19 pandemic that I assembled from the state-run news agency and a progovernment news outlet, this article explores how Turkey’s foreign aid policies have shaped the distribution of Turkish medical aid across countries during the pandemic. The results of logistic regression analysis show that being a recipient of Turkey’s medical aid is associated with the recipient countries’ historical, cultural, and economic relations with Turkey. In particular, having historical ties to the Ottoman Empire is the most substantive indicator of receiving Turkish medical aid. With regard to cultural relations, ethnic affinity is a strong predictor of receiving Turkish medical aid, whereas having a common religion does not show a significant relationship. Therefore, this article demonstrates that previous policies of the Turkish state with regard to the provision of foreign aid are largely at play in the case of the coronavirus pandemic.

By conducting statistical analysis on the TurAid dataset, the findings of the article have important empirical implications for recent theoretical discussions about the determinants of Turkish foreign aid policy (Mehmetcik and Pekel 2020; Kavaklı 2018; Zengin and Korkmaz 2019; Ipek 2015; Altunışık 2014). As the Turkish state has hitherto considered a number of different factors in determining where to send foreign aid, the COVID-19 pandemic has given us an important opportunity to look at these established theoretical outlooks in the light of new evidence. This may lead to rethink how Turkish foreign aid policy may change in the event of an unexpected worldwide crisis. In this respect, the article makes two contributions to the literature on the determinants of Turkish foreign aid. It is the first systematic analysis of Turkish medical aid during the COVID-19 pandemic. Equally important, it also utilizes a novel media-based event dataset. All in all, this article investigates to what extent Turkey’s foreign aid policy influences the spatial variation of Turkish medical aid in managing an exceptional global crisis.

This article is structured around four main sections. The next section presents an overview of the TurAid dataset, including giving detailed information about the data collection process as well as the variables. The third section accounts for Turkey’s selective strategy in sending medical aid across countries. By making use of statistical analysis, this section explains the spatial variation of Turkish medical aid on the basis of Turkey’s foreign aid policy. I summarize the main findings, limitations, and future avenues of research in the conclusion.

The TurAid dataset: a comprehensive overview

I assembled a novel dataset concerning Turkey’s global medical aid (TurAid) during the COVID-19 pandemic between February 1^st and July 31^st, 2020. The data was collected from online news reported by Turkey’s state-run news agency, Anadolu Agency (AA), and a pro-government news outlet, Daily Sabah (DS). I chose AA and DS because they are the only news organizations that cover and keep updated all state and government-related events in a timely manner. Furthermore, they have emerged as the conduits of the Turkish state’s public diplomacy in the past few years (Akşak 2020; Çevik 2018). While AA reports in more than 10 different languages, DS only reports in English and Arabic. Turkey’s cross-border operations in Syria, the dispute over the exclusive economic zone between Turkey and Greece in the Eastern Mediterranean, and Turkey’s indirect intervention in the Libyan Civil War are some recent examples in which both news sources have maintained extensive coverage. Therefore, they stand out as important media to capture data on the supply of Turkish medical aid during the coronavirus pandemic.

A news search produced 108 unique items of news, 15 photo albums, 10 videos, and 1 infographic on the website of AA. 105 out of the 108 news items contained content relating to Turkey’s global efforts in providing medical aid. Only three items were identified as not relevant. This 97% rate of precision indicates the percentage of congruence between tags in each item of news and the keywords searched. Moreover, 61 different countries were captured as a result of the keyword search on AA. Regarding DS, a news search produced 272 unique items of news. 123 out of the 272 contained relevant content, a 45% rate of precision. Given these percentages, it is apparent that DS generally uses more irrelevant tags for its news items than AA, resulting in a lower level of matching. Only 7 countries were unique to DS: Brazil, Ireland, France, Qatar, Togo, Haiti, and Mexico.

To make sure that the dataset was not missing any countries receiving Turkish medical aid, I also checked news featured on the websites of domestic donor institutions. These institutions are TIKA (Turkish Cooperation and Development Agency), Turkish Red Crescent, and the Ministry of Health. After having carried out the same keyword search, only TIKA listed additional countries: Eswatini, Burkina Faso, Cameroon, and Botswana. In this respect, the dataset currently covers 184 online news items from 72 countries with 30 variables between February 1^st and July 31^st, 2020. I began in February 2020 as it was the first time that Turkey sent medical supplies to a country owing to the COVID-19 pandemic – the city of Wuhan in China.

The news extraction process was composed of the following steps: (i) visiting the English-version websites of AA and DS; (ii) going to the search bar on the website; (iii) entering the relevant keywords ‘medical’, ‘aid’, ‘Turkey’; (iv) selecting the appropriate news category and date ranges with respect to the publication dates of each news report. The unit of analysis was an event-day-location which was a medical aid event reported on the websites of AA and DS on a certain day in a particular location. The dataset has 467 medical aid events. In respect of the coding of these aid events, three caveats should be made. Firstly, medical aid events in different countries appearing in the same item of news are coded separately.¹ Next, multiple items about one country appearing on the same day are also coded separately.² Lastly, some news items include medical aid sent to certain international as well as regional organizations, such as NATO. If there was no specific information in the news about which of the countries were recipients of aid from such organizations, I coded them as ‘Regional’ or ‘Global’ according to their contents.³

This study employs content analysis of the news published on the websites of AA and DS concerning Turkish medical aid events during the pandemic. Six different news categories appear in the dataset.⁴ Most news items belong to the multiple entry category (184), constituting 40% of the total. This is followed by single entry (93), praise (72), other (62), global (28) and regional cooperation (28), respectively. The number of aid events categorized as ‘other’ should not be misunderstood. Most of the aid events in this category come from a very comprehensive infographic in which AA shows more than 40 different countries to whom Turkey provided medical aid.⁵ The dataset also includes the full names and locations of each correspondent reporting the news.

In addition to the textual parts of the news, I also coded the images used, grouping them into six categories. Certain images are inserted into multiple news items. The categories are: aid delivery (143), loading/unloading aid to/from a vehicle (99), political and diplomatic officials (84), other (53), multiple (50), and aid preparation (38). News editors select photos that demonstrably illustrate Turkey’s efforts in providing medical supplies. Moreover, most news items contain four motifs in common: the emblem of the Presidency; the Turkish flag; the emblems of the organizations providing medical aid, such as the Turkish Red Crescent; and the Rumi quote mentioned at the beginning of this article. These motifs are embedded into medical aid packages.

The contents of the medical aid are specified in around 60% of aid events (266 out of 467 aid events).⁶ 210 aid events contain protective equipment. This protective equipment includes N95 masks, surgical masks and gowns, face shields and vizors, protective clothing, glasses, gloves, overalls, and goggles. The focus on the supply of protective equipment is not surprising since the ultimate aim, at the first stage, was to reduce the high numbers of infected individuals in many countries (BBC News 2020b). Considering the huge demand and increasing price of protective materials, the supply of protective equipment globally quickly emerged as an important issue.⁷ Protective equipment is, respectively, followed by testing kits (48), medical devices (34), others (30), nutrition (26), and medical drugs (8). The ‘others’ category comprises the renovation and building of hospitals, hospital beds, equipment for intensive care units, and medical vehicles, such as ambulances.

By the same token, diagnostic materials were also important to identify and thereby trace the movement of patients. To this end, Turkey has provided recipients with testing kits and PCR machines. In this respect, it should be noted that the Turkish state in partnership with the private sector started to produce breathing devices in late March 2020. While the country was providing these machines for national use, it also supplied these domestically-produced mechanical ventilators to countries that could not meet the heavy demands of patients particularly in intensive care units.⁸

TurAid has another variable to understand whether a specific donor institution is indicated in the aid events. Around 25% of the total aid events (120 out of 467) specify the name of the Turkish state and private institution providing the aid. TIKA’s name appears on nearly 70% of the aid events (81 out of 120) where the donor is known. Considering the active role that TIKA has played generally in providing world-wide development and humanitarian assistance in the last decade, this is not surprising.⁹ It is followed by the Ministry of Health (27), other public or private organizations such as the Turkish Exporters’ Assembly (7), and the Turkish Red Crescent (5).

The delivery of medical aid establishes direct political ties between donor and recipient countries (Alesina and Dollar 2000). This is why it is important to explore whether the aid events cover recipient countries expressing gratitude to Turkey. To do that, I created a dichotomous variable for the appreciation of medical aid by the aid-receiver. From the perspective of Turkey as the donor country, the news could also explicitly promote its efforts to supply global medical aid by sharing information about previous batches of medical aid sent to other countries. While recipient countries thank Turkey for the medical aid in 133 aid events, Turkey praises its own efforts in supplying global medical equipment in a total of 351 aid events. Moreover, 97 aid events entail both of the above-mentioned categories. In addition, as the rapid spread of the COVID-19 pandemic led to a sudden cancellation of international flights, many people were left stranded in foreign countries (World Economic Forum 2020). For this reason, I coded whether the news mentions the repatriation of individuals from the recipient country as well as Turkey.

Deciphering Turkey’s selective provision of medical aid during the COVID-19 pandemic

Out of 467 aid events, 72 countries emerge as aid-recipients. Figure 1 depicts the distribution of Turkish medical aid around the world. Turkey has sent medical aid to at least one country in every continent, except for Australia and Antarctica. Cross-continent variation shows that Asia and Africa enjoyed more Turkish medical aid than others, with 22 countries in each continent. They are closely followed by Europe, with 20 countries. On the other hand, the Americas seem underrepresented. While 5 countries in South America have received medical aid from Turkey, only 3 countries in North America have done so: the United States (US), Mexico, and Haiti. The TurAid dataset also shows that only 27 cities are specifically referred to in the news. Most of these cities are in Muslim-majority countries, or countries with a considerable Muslim population. For instance, Turkey has sent multiple batches of medical aid to Novi Pazar, which is a Muslim-majority city in Serbia (Anadolu Agency 2020b).

Figure 1. Distribution of Turkish medical aid across countries during the pandemic

So, what explains the spatial distribution of Turkish medical aid during the coronavirus pandemic? I argue that Turkey’s selective strategy in the supply of global medical aid is related to the three determinants of Turkey’s previous aid behaviours. These are based on Turkey’s historical, cultural (including ethnic and religious affinities), and economic relationships with other countries. This section reports statistical analysis of the spatial variation of Turkish medical aid during the pandemic; furthermore, it discusses the findings based on the determinants of Turkish foreign aid policy. The dependent variable is binary, as it looks for whether or not a country received Turkey’s medical aid during the pandemic.¹⁰

Independent variables

The ‘Neo-Ottomanist’ foreign policy of the Turkish government requires it to be responsible for regions which were part of the Ottoman Empire for centuries (Murinson 2006, 21). Equally important, the zero-problems with neighbours policy of Ahmet Davutoğlu, a former foreign and prime minister of Turkey in past governments, has played a crucial role in decisions to send development and humanitarian aid to neighbouring countries in the past decade (Davutoğlu 2010). Thus, the first indicator for receiving Turkish medical aid relates to a country’s historical ties to the Ottoman Empire. To this end, I created a binary variable as to whether countries fully or partly had belonged to the Ottoman Empire.

‘Cultural commonality’ is an important determinant of Turkish foreign aid behaviour (Altunışık 2014, 336). The first component of cultural similarity is Turkey’s religious affinity with recipient countries (Kavaklı 2018; Fisher Onar 2011). The Muslim world has received significant attention from Turkey since the early 2000s. More specifically, Turkish foreign aid has focused on Muslim communities with different types of humanitarian and poverty-related problems. For example, the Turkish state has sent different batches of humanitarian aid to the Rohingya Muslims since 2015 owing to their forcible displacement from the Arrakhan and Rakhine states of Myanmar to neighbouring countries – Bangladesh in particular. Somalia is another example of Islam-motivated Turkish foreign aid, in which Turkey has given a considerable amount of development and humanitarian aid in the past decade (Akpınar 2013). At this point, I posit that it is likely that Turkey, as a Muslim-majority country, would provide Muslim countries and countries comprising significant Muslim minority populations with medical aid during the coronavirus pandemic in the name of Islamic solidarity. To investigate the aforementioned hypothesis, another indicator of receiving Turkish medical aid is the percentages of the Muslim population in recipient countries (Maoz and Henderson 2013).¹¹

The other aspect of cultural similarity that the paper examines is the Turkish state’s ethnic sensibilities. When we look at Turkey’s previous efforts in giving foreign aid, countries with a Turkish population have largely benefitted from Turkish development and humanitarian aid (Çevik and Seib 2015; Altunışık 2014; Köstem 2017; Bilgin and Bilgiç 2011). To explore whether these ethnic sensitivities hold in the case of the pandemic, I operationalize this ethnic brotherhood using the percentages of Turkic languages spoken in countries receiving aid (Fearon 2003). By looking at the impact of a common language between Turkey and other countries, we will be able to understand whether ethnicity is a factor in receiving medical aid from Turkey.

Economic determinants of Turkish foreign aid are two-pronged. Firstly, foreign aid can be utilized for ‘ … strengthening economic ties between the donor and the recipient.’ (Mehmetcik and Pekel 2020, 139). In this respect, Turkey has been able to make use of medical aid during the pandemic as an important gesture to countries with whom it maintains varying levels of economic relations. To this end, Turkey’s economic relations with countries are operationalized by the 2019 export share of Turkey (logged) in recipient countries (Türkiye Istatistik Kurumu [the Turkish Statistical Institute] 2020). Regardless of their bilateral economic relations, Turkey also considers potential aid-recipient’s economic capabilities in handling various crises (Korkmaz and Zengin 2020). Considering the characteristics of the COVID-19 crisis, the recipients’ needs based on health system performance is an important indicator in understanding the resilience of countries facing such a devastating public health crisis (European Commission 2017). I capture recipients’ needs through infant mortality rate, which is defined as the number of infants dying prior to their first birthday per 1,000 live births, as an indicator for ‘the effectiveness of national health system’ (World Health Organization 2020, 44), as well as poverty (World Bank 2019b).¹² Lastly, control variables include countries’ populations (World Bank 2019c) and distance to Turkey (Mayer and Zignago 2011).¹³ Both variables are logged.¹⁴

Results

A logistic regression analysis helps to explore how three different determinants of Turkish foreign aid policy can predict whether a country benefits from Turkey’s medical aid during the pandemic.¹⁵ The results show the following: while the determinants related to economic and historical ties to Turkey are supported by the data, we have partial evidence for cultural determinants in respect of being a recipient of Turkish medical aid. There is a statistically significant relationship between Turkic language spoken in countries and receiving Turkish medical aid. However, the percentage of the Muslim population in a country does not seem to be a predictor of benefitting from Turkey’s medical aid. As for the control variables, while having Turkish medical aid is not statistically associated with distance to Turkey, a country’s population is positively and significantly associated with receiving aid. Table 1Table 2 demonstrates summary statistics of each variable used in the model.

Table 1. Logistic regression of receiving Turkish medical aid during the pandemic

Variables	Odd ratios
Historical ties to the Ottoman Empire	5.285**
	(4.286)
Turkic language spoken	2.233*
	(0.924)
Muslim population	0.995
	(0.006)
Export share (logged)	2.217**
	(0.753)
Infant mortality rate	1.040***
	(0.013)
Distance (logged)	0.466
	(0.378)
Population (logged)	2.027**
	(0.727)
Constant	0.0002**
	(0.001)
Observations	177

R2 = .359 (Cox & Snell), .489 (Nagelkerke).

Standard errors are in parentheses.

* p<0.1, ** p<0.05, *** p<0.01

Table 2. Summary statistics

Variables	N	Mean	SD	Min	Max
Medical aid	196	0.367	0.483	0	1
Historical ties to the Ottoman Empire	196	0.117	0.323	0	1
Turkic language spoken	196	2.482	12.44	0	90
Muslim population	188	24.98	36.00	0	99.56
Export share (logged)	189	5.045	1.160	1.28	7.22
Infant mortality rate	192	21.45	19.46	1.4	84.5
Distance (logged)	185	3.657	0.368	2.645	4.236
Population (logged)	196	6.743	1.006	2.903	9.145

Table 1 indicates that statistically significant variables show different levels of substantiveness based on their odds ratio. For example, the odds of being the recipient of Turkish medical aid for countries which fully or partially belonged to the Ottoman Empire, over the odds of getting medical aid for countries without an Ottoman past, is 5.285. In other words, we can say that the odds for countries with an Ottoman past are 428,5% higher than the odds for other countries. Similarly, for a one percent increase in the rate of Turkic language spoken, we see a 123.3% increase in the odds of receiving Turkish medical aid. With regard to economic motivations, we see a 121.7% increase in the odds of receiving Turkish medical aid for every one unit increase in Turkey’s export share in a country, as the odds ratio is 2.217. Finally, for a one percent increase in the infant mortality rate, we see a 4% increase in the odds of benefitting from Turkish medical aid.

Discussion

Similar to the findings of Kavaklı (2018), Zengin and Korkmaz (2019), and Korkmaz and Zengin (2020), countries with an Ottoman past have been significantly more likely to receive Turkish medical aid during the pandemic. Out of 23 countries that partially or fully belonged to the Ottoman Empire, we see only four countries which did not enjoy Turkish medical aid: Armenia, Egypt, Greece, and Slovakia. Turkey’s current and established relations with most of these countries would explain the absence of medical aid. Considering the tumultuous relations between Greece and Turkey, the absence of medical aid to Greece might not be that surprising. Contemporary relations between the countries show different dimensions of disagreement in the past few years. While we can still observe issues that are historically grounded, such as the Ecumenical Patriarchate of Istanbul and the Turks of Western Thrace, both countries have also clashed over more recent topics, such as problems germane to natural gas exploration in the Eastern Mediterranean.

Similarly, Armenia-Turkey relations also comprise contemporary and historical issues, such as the dispute over the recognition and nomenclature of the ‘Armenian Genocide’. Although President Erdogan has announced that Turkey commiserates with Armenian society regarding ‘the events of 1915ʹ, the Turkish state does not recognize those events as a genocide (BBC News 2014). This has been a longstanding controversy between both countries, leading to a strong reluctance to revitalizing bilateral diplomatic relations. Looking at contemporary issues, Turkey has been vociferous in its criticism of Armenia’s armed intervention in the disputed Nagorno-Karabakh region between Armenia and Azerbaijan (German 2012). The ruling government of Turkey sides with Azerbaijan in protecting Azeri interests in the region. The Turkish parliament’s attitude towards the 2020 Armenian-Azerbaijani conflict is a clear example of Turkey’s position in the dispute (Anadolu Agency 2020c). Turkey’s relations with Egypt have been tense since the 2013 coup d’état in Egypt. Both countries expelled each other’s ambassadors and diplomatic relations were downgraded shortly after the coup (BBC News 2013). Recently, they have been at odds over the agreement that Egypt has recently signed with Greece with regards to an exclusive economic zone in the Eastern Mediterranean (Reuters 2020).

Noteworthily, Cyprus and Syria have also received medical aid; however, this aid has been delivered to the northern parts of both these countries.¹⁶ In other words, the aid has not been sent to the state authorities of these countries. Emanating from its long-running dispute with Cyprus, Turkey has no diplomatic relations with it. Nevertheless, the Turkish Republic of Northern Cyprus (TRNC), which is a de facto state, has received Turkish medical aid, Turkey being the only country which recognizes the TRNC. Similarly, since the beginning of the Syrian Civil War, Turkey and Syria have had a highly fractious relationship, leading to a number of military operations in the past years. Notwithstanding this, Turkey has dispatched multiple batches of medical aid to the northwestern part of Syria, where we observe high levels of control by the Turkish Armed Forces and the Syrian Free Army.

Among the countries covered by the TurAid dataset, 27 recipients have a Muslim-majority population where individuals identifying as believers of Islam constitute more than 50% of the total population. In addition, 5 countries have considerable Muslim minorities, ranging from 20 to 40% of the total population: Ethiopia, North Macedonia, Cote d’Ivoire, Cameroon, and Mozambique. Almost half of the countries covered by the TurAid dataset come from these two groups of countries.¹⁷ When we look at previous studies, we see that they do not have consistent findings with regard to the effect of common religion on the allocation of Turkish foreign aid. Mehmetcik and Pekel (2020) show a significant relationship between the percentage of the Muslim population in recipient countries and receiving Turkish foreign aid. More specifically, Kavaklı (2018) finds that the effect of a common religion increases during the period of AKP rule from 2003. On the other hand, while Zengin and Korkmaz (2019) demonstrate the significance of Islam in Turkish foreign aid, this significance is reduced in former Ottoman lands and Turkic republics compared with other countries.¹⁸ In line with my findings, Korkmaz and Zengin (2020) do not find a statistically significant role for Islam in allocating Turkish foreign aid.¹⁹

At this point, an important question arises as to why Turkey did not overwhelmingly concentrate its efforts on supplying medical aid to the Muslim world. I argue that the answer lies behind Turkey’s aspiration to exert its soft power in countries where it wants to shape the perceptions of the general public towards Turkey. In other words, I claim that Turkish foreign aid policy has been used as an important instrument of Turkey’s soft power during the pandemic. If Turkey had specifically focused on the Muslim world in sending medical aid, it would create an impression in non-Muslim countries that the Turkish state only meets the needs and demands of its co-religionists.

The Soft Power Index (SPI) can give us more nuanced information about Turkey’s soft power (McClory 2019). The SPI annually ranks the Top 30 countries according to their soft power scores since 2015. Turkey’s presence in the ranking varies across years.²⁰ The SPI is composed of a variety of sub-indexes. The ‘digital’ and ‘engagement’ aspects of soft power have played a decisive role in promoting Turkey onto the list. While the engagement sub-index has metrics like ‘the number of embassies/high commissions a country has abroad, membership of multilateral organizations, and overseas development aid contributions’, the digital sub-index ‘ … captures the extent to which countries have embraced technology, how well they are connected to the digital world, and their use of digital diplomacy through social media platforms’ (McClory 2019, 27–28).

Parallel to the engagement sub-index, President Erdogan has paid a considerable number of work and state visits to various countries in different continents since taking office. For instance, as President, Erdogan has visited 7 countries in South America and 23 countries in Africa (Türkiye Cumhuriyeti Cumhurbaskanligi [Presidency of the Republic of Turkey] 2020) .²¹ In addition, the increasing number of diplomatic missions in both continents provides us a vivid view of Turkey’s diplomatic efforts in those regions. While Turkey had diplomatic missions in only 12 African countries in 2002, this has increased to 42 countries to date (Anadolu Agency 2020a). In respect of South America, Turkey has opened 5 new embassies in the last decade, whereas the country had only 5 embassies until 2010.²² Concerning the digital sub-index, the publication of news covering Turkey’s medical aid on the social media accounts of AA and DS also show how effectively Turkey uses the digital sphere for its public diplomacy.

Apart from Mehmetcik and Pekel (2020), previous research finds that ethnic affinity is significantly associated with the provision of Turkish foreign aid (Kavaklı 2018; Korkmaz and Zengin 2020; Zengin and Korkmaz 2019). Notwithstanding the aforementioned, Kavaklı (2018) and Zengin and Korkmaz (2019) demonstrate that the effect of a common language in Turkish foreign aid has decreased during AKP rule. Hence, the results of this article are consistent with the extant literature because the percentage of a Turkic language spoken in countries is statistically associated with being a recipient of Turkish medical aid. In this respect, the Turkic Council is an important institution to understand Turkey’s ethnic sensitivity. The council has 5 members and 1 observer: Azerbaijan, Kazakhstan, Kyrgyzstan, Uzbekistan, Turkey, with Hungary as the observer state. The TurAid dataset shows that Turkey has sent medical aid to all of the Council countries except Turkmenistan. The reason why Turkmenistan is not included in the dataset might be two-fold. Either the country may not have suffered severely from a lack of medical equipment or might have limited access to COVID-19-related information. The former might be plausible as the country has reported no cases of COVID-19 thus far (Human Rights Watch 2020; BBC News 2020a). However, considering the repressive characteristics of the Turkmen state about information sharing, the coronavirus data in Turkmenistan may not be reliable.²³

All the quantitative studies of Turkish foreign aid find that the role of Turkish exports is significantly related with sending foreign aid (Kavaklı 2018; Korkmaz and Zengin 2020; Zengin and Korkmaz 2019; Mehmetcik and Pekel 2020). In this regard, this research supports these prior studies. This is not surprising when we look at Turkey’s top 20 export partners in the last decade. 15 out of the top 20 export partners received Turkish medical aid during the pandemic.²⁴ This supports the idea that the Turkish state has been quite responsive with its close export partners, ensuring that its export relations are not adversely affected by any unresponsiveness on its part during the pandemic. Nonetheless, the significant relationship between recipients’ needs and receiving Turkish medical aid shows that Turkey also seems to have been concerned with the economic capacities of countries. For example, 15 out of the top 30 worst countries in terms of infant mortality rate, ranging between 84.5 (Central African Republic) and 42.90 (Yemen) have received Turkish medical aid.

With regard to the extant literature, Kavaklı (2018) finds that infant mortality rate is negatively and significantly related to Turkey’s humanitarian aid. In the case of Turkish medical aid, a positive and significant relationship can be associated with the urgency of the crisis. In other words, any intervention needed to be deployed quickly by recipients to control the spread of the pandemic, irrespective of their logistical infrastructure. In a similar way to the measure of infant mortality rate that I use, Mehmetcik and Pekel (2020) employ life expectancy at birth to understand the humanitarian aspect of Turkish foreign aid. They find a negative and significant association between life expectancy and such aid.

In the light of the significant impact of recipients’ needs in respect of the provision of Turkish foreign aid, I argue that Turkey maintains a delicate balance in determining its aid-recipients. Put differently, while Turkey shows solidarity with its export partners, it also seeks to respond to countries in need of assistance. This has some important implications about how we conceptualize Turkey’s foreign policy vision. Turkey has particularly attached importance to humanitarian diplomacy over the last decade, linking power with humanitarian causes and conscience-led policies (Davutoğlu 2013). To this end, it has shifted some of its attention to regions suffering severe humanitarian crises, through different forms of aid. On the other hand, while Turkey has maintained export activities with established trade partners such as Germany and France, it has also been able to explore new markets as a result of its increasing diplomatic and political relations in different countries. Given the fact that Turkey’s GDP per capita has been declining since 2013 (World Bank 2019a), the statistically significant and positive impact of Turkish exports on decisions to send foreign aid can be conceived as an important gesture in favour of its trade partners which may make a positive contribution to Turkey’s future export performance. Beneath Turkey’s long-standing aid strategy, there is a clear interplay between economic pragmatism and humanitarianism.

Conclusion

Why have some countries received medical aid from Turkey during the COVID-19 pandemic but not others? Relying on a novel dataset that has been assembled from the state-run news agency and a pro-government news outlet, this article explores whether Turkish medical aid during the coronavirus pandemic shows similar patterns with the determinants of Turkey’s foreign aid policy without the role of common religion. Logistic regression analysis of the receipt of Turkish medical aid shows that there is a statistically significant relationship between such aid and recipient countries’ historical, cultural as well as economic relations with Turkey during the COVID-19 pandemic. Specifically, with regard to cultural relations between Turkey and recipient countries, we do not see any statistically significant relationship between Islamic affinity and being a medical aid-recipient, even though ethnic kinship indicates a significant association.

Considering the results of this paper, I argue that the theoretical foundations of Turkey’s general policies in delivering foreign aid apply in the case of the COVID-19 pandemic, except that there is only partial evidence of the importance of cultural relations. Furthermore, drawing on a unique dataset, this article is the first systematic analysis of the spatial variation of Turkey’s distribution of global medical aid during the pandemic. The findings of this article connect Turkey’s general foreign aid behaviour with its specific efforts to give medical aid during the pandemic. As a result, this research demonstrates that Turkey has adopted a selective aid strategy for an indiscriminately damaging global crisis.

The possible limitations of this article relate to the data collection process. It might be the case that the retrieval of news from AA and DS could lead to both under- and over-coverage biases. As shown in the overview of the TurAid dataset, these news outlets stand out as important tools of Turkey’s public diplomacy. Thus, if Turkey sends medical aid to any country, it is highly likely that one or both news platforms would cover it to contribute to Turkey’s positive image in international politics. This mitigates the possibility of under-coverage. Moreover, as the news is reported in English, there would likely be a backlash from recipients if they published any misinformation about the delivery of medical aid to such countries. This means that both agencies have an important responsibility to give reliable information about such medical aid events. Therefore, it would make over-coverage highly costly for their editorial boards.

Future research may want to focus on why certain countries received multiple aid packages in comparison to others, or why some countries specifically received certain types of medical aid. This kind of research would augment the literature on foreign aid. Additionally, the political communication aspects of Turkish medical aid may also be a fruitful topic of research. It would also be relevant to ask whether news covering Turkish medical aid is differently tailored for domestic and international audiences. In this respect, it may be seen that the promotion of Turkey’s global medical aid in Turkish brings out different patterns than its promotion in English. These two strands of future research would certainly illuminate studies about the impacts of the COVID-19 pandemic on the supply of Turkey’s medical aid.

Acknowledgments

I would like to thank Kerim Can Kavaklı and two anonymous referees for useful feedback.

Supplementary material

Supplemental data for this article can be accessed here.

Disclosure statement

No potential conflict of interest was reported by the author.

Data availability statement

Data and code for this article are available at https://doi.org/10.7910/DVN/KDIEGV.

Additional information

Notes on contributors

Buğra Güngör

Buğra Güngör is a Ph.D. candidate in the Department of International Relations and Political Science at the Graduate Institute of International and Development Studies in Geneva. He also works as a Teaching Assistant for Capstone Projects in the Development Studies programme at the same institution. He completed his MA in Conflict Analysis and Resolution at Sabancı University and his BA in Philosophy at Bilkent University. His research focuses on conflict processes, public diplomacy, soft power, Turkish foreign policy, and the EU-Turkey relations.

Notes

1. For instance, medical aid sent to the Balkan countries are often covered in one news item. As an example, please see the following: Daily Sabah (2020) ‘Turkey contributes to Balkans’ fight against coronavirus with medical aid, solidarity’, 08 April, available online at: https://www.dailysabah.com/politics/diplomacy/turkey-contributes-to-balkans-fight-against-coronavirus-with-medical-aid-solidarity.

2. For example, medical aid sent to Kosovo appears in two different news items on Anadolu Agency. They also denote different types of news as ‘praise’ (i) and ‘multiple’ (ii) categories. The news respectively follows: (i) Kosovo president thanks Turkey for help in virus fight: (i) Anadolu Agency (2020) ‘We are deeply grateful to Turkey for medical aid, says Hashim Thaci’, 08 April, available online at: https://www.aa.com.tr/en/europe/kosovo-president-thanks-turkey-for-help-in-virus-fight/1797301 and (ii) Anadolu Agency (2020) ‘ ‘Turkey’s helping hand to be always on our brothers’: Turkey will never leave Balkan countries alone in their fight against coronavirus, says parliament speaker’ 08 April, online available at: https://www.aa.com.tr/en/turkey/turkeys-helping-hand-to-be-always-on-our-brothers/1797237.

3. The following news is an example: Anadolu Agency (2020) ‘Turkey distributes protective supplies to NATO staff: Defence Ministry provides medical equipment to staff of 23 NATO member countries stationed at base in western Turkey’, 15 May, available online at: https://www.aa.com.tr/en/latest-on-coronavirus-outbreak/turkey-distributes-protective-supplies-to-nato-staff/1842485.

4. Please refer to Table A.2 and Table A.3 in the Supplementary Materials to see the coding schemes for the categories of news and images used.

5. Please find the relevant infographic in the following news: Anadolu Agency (2020) ‘Turkey sends aid to at least 57 countries to fight virus: While fighting the virus domestically with weekend lockdowns, quarantines and social isolation, Turkey continues to supply medical aid to many countries’, 28 April, online available at: https://www.aa.com.tr/en/info/infografik/18383.

6. Some aid events contain multiple types of medical aid. Please see the following news report by way of example: Daily Sabah (2020) ‘Turkey to send aid to Niger, Chad to fight COVID-19 pandemic’, 17 June, available online at: https://www.dailysabah.com/politics/diplomacy/turkey-to-send-aid-to-niger-chad-to-fight-covid-19-pandemic.

7. Tedros Adhanom Ghebreyesus, who is the current Director General of the World Health Organization, made some remarks at a media briefing about the coronavirus. He said that the prices of protective equipment had increased six times and there was limited access to certain supplies like medical masks. For further information, please see the following link: World Health Organization (2020) ‘WHO Director-General’s opening remarks at the media briefing on COVID-19 – 3 March 2020ʹ, 3 March, available online at: https://www.who.int/dg/speeches/detail/who-director-general-s-opening-remarks-at-the-media-briefing-on-covid-19—3-march-2020.

8. As an example of the supply of domestically-produced ventilators, Brazil has received more than 600 mechanical ventilators to increase the capacity of intensive care units. Please see the following news: Daily Sabah (2020) ‘Turkey’s ventilators help Brazil combat COVID-19 pandemic’, 19 June, available online at: https://www.dailysabah.com/politics/diplomacy/turkeys-ventilators-help-brazil-combat-covid-19-pandemic.

9. Serdar Cam, who is the previous president of TIKA, gave an interview about the role of TIKA in Turkey’s humanitarian aid and development assistance. Please see the following news: Daily Sabah (2018) ‘TIKA President Serdar Cam: Turkey keeps breaking records in humanitarian aid and development assistance’, 10 December, available online at: https://www.dailysabah.com/politics/2018/12/10/tika-president-serdar-cam-turkey-keeps-breaking-records-in-humanitarian-aid-and-development-assistance.

10. Countries are selected according to the UN member (193) and observer states (The Holy See and the State of Palestine). Given the complicated situation of its international recognition, we also have Kosovo. Therefore, there are 196 countries in total.

11. The insignificance of Islamic identity persists if we use a dichotomous measure of religious commonality based on the largest religious group in a recipient – please see Table A.15 in the Supplementary Materials.

12. Consistent with Kavaklı (2018), I added gross domestic product (GDP) to the model to control for the recipients’ level of income along with the infant mortality rate as a related, but distinct, indicator of the recipients’ needs (World Bank 2019a). The GDP did not turn out to be a predictor of receiving medical aid; what is more, its inclusion did not change statistically significant independent variables in the model. Please see Table A.16 in the Supplementary Materials.

13. The data on distance is calculated according to the geodesic distances between Turkey and each country which looks for latitudes and longitudes of the most significant cities/regions in terms of population.

14. Collinearity diagnostics demonstrate that none of the independent variables can be suspected of high linearity. Please see Table A.13 in the Supplementary Materials.

15. Please refer to Figure A.2–4 and Table A.7–10 in the Supplementary Materials to see the post-estimation statistics.

16. Statistically significant variables do not change when we remove Syria and Cyprus from the logistic regression anaylsis. Please see Table A.11 in the Supplementary Materials.

17. Please refer to Table A.4 in the Supplementary Materials to see the list of Muslim-majority countries and the countries with significant Muslim minorities in the TurAid dataset.

18. Relying on Zengin and Korkmaz (2019), interactions between (i) countries with an Ottoman past and the percentage of Muslim population and (ii) the percentage of Turkic language spoken and the ratio of Muslim population do not show a statistically significant association in receiving Turkish medical aid. The insignificance of the percentage of Muslim population in having Turkish foreign aid remains the same when we include two interaction variables, whereas other independent variables do not keep their significance. This means that the model is not valid for this set-up. Please refer to Table A.12 in the Supplementary Materials.

19. To ensure that Islamic identity and Turkic identity do not overlap with each other, I run separate regression analysis with each of these variables. The insignificance of Islamic identity persists when I exclude Turkic identity, whereas Turkic identity shows a significant and positive association if we remove Islamic identity. Therefore, this gives me the same results that I find in the full model. Please refer to Table A.14 in the Supplementary Materials.

20. Turkey has entered the Top 30 in 2015 (28^th), 2017 (30^th), and 2019 (29^th) since the release of the first ranking in 2015.

21. Please refer to Table A.6 in the Supplementary Materials to see the list of Erdogan’s presidential visits in South America and Africa between September 2014 – January 2020.

22. Recently opened embassies are located in Colombia (2010), Peru (2010), Ecuador (2012), Bolivia (2018), and Paraguay (2018).

23. Similar to the specific cases of Syria and Cyprus, Turkmenistan and North Korea are removed from the logistic regression anaylsis. Statistically significant variables remain the same. Please see Table A.11 in the Supplementary Materials.

24. Please refer to Table A.5 in the Supplementary Materials to see the list of Turkey’s top 20 export partners between 2009 and 2018 and whether they received Turkish medical aid.