Research, Methods and Analysis in Social Sciences and Humanities 2025 – III

Research, Methods and Analysis in Social Sciences and Humanities 2025 - III Editor Gülsen KIRAL Lyon 2025 Research, Methods and Analysis in Social Sciences and Humanities 2025 - III Editor Gülsen KIRAL Lyon 2025 Research, Methods and Analysis in Social Sciences and Humanities 2025 – III Editor • Prof. Dr. Gülsen KIRAL • Orcid: 0000-0002-0541-0178 Cover Design • Motion Graphics Book Layout • Motion Graphics First Published • October 2025, Lyon e-ISBN: 978-2-38236-945-6 DOI: 10.5281/zenodo.17436709 copyright © 2025 by Livre de Lyon All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without prior written permission from the Publisher. The author or authors of the relevant section are responsible for any copyright infringement that may occur due to the images and graphics used in the book. The editor or publisher does not assume responsibility in this regard. Publisher • Livre de Lyon Address • 37 rue marietton, 69009, Lyon France website • http://www.livredelyon.com e-mail • [email protected] I PREFACE The book, titled “Research, Methods and Analysis in Social Sciences and Humanities" covers research undertaken by highly regarded academic members from various parts of Turkey. The subject involves interdisciplinary research in areas that include Business Administration, Finance, Econometrics, and Political Science. There are nine chapters in the book The book examines into both theoretical and applied approaches that are often used in the discipline of Social sciences. The book covers a range of current subjects, such as monetary policies, survival analysis, management strategies, banking system, artificial intellience, public finance. Also you can find two different Panel data analysis applications which are examined the Word Bank data sets. We are certain that this book will offer assistance to a wide range of readers who have an intense curiosity in studying and carrying out research in the field of social sciences. We believe that our book will be of great assistance to both the instructors and graduate students who have an active role in social sciences research. Additionally, we expect that it will also be beneficial to a broader spectrum of readers. I would like to express my appreciation to all the authors who willingly offered their valuable time to contributing to our book, as well as to all the publishing company staff who carefully and professionally participated in numerous responsibilities, including formatting and printing. Prof. Dr. Gülsen KIRAL III İÇİNDEKİLER PREFACE I CHAPTER I. THE IMPACT OF THE 1925 FRIENDSHIP AND NEUTRALITY TREATY ON TURKEY-USSR RELATIONS 1 Abdullah TORUN CHAPTER II. AN ANALYSIS OF THE 44-DAY KARABAKH VICTORY IN THE CONTEXT OF “TWO STATES ONE NATION” IN TÜRKİYE–AZERBAIJAN RELATIONS 15 Elnur Hasan MİKAİL & Hakan ÇORA CHAPTER III. THE MONETARY POLICY OF THE EUROPEAN CENTRAL BANK: A LITERATURE REVIEW 33 Mehmet Emre ÜNSAL CHAPTER IV. UNEMPLOYMENT AND EDUCATION EXPENDITURES: TESTING FOR CAUSALITY USING DUMITRESCUHURLIN APPROACH 49 Gülsüm GÜRLER HAZMAN & Almina Derya BIÇAKSIZ CHAPTER V. HOW TO DEAL WITH MISSING OBSERVATIONS IN 67 PANEL DATA ANALYSIS USING STATA Cansu UNVER ERBAS CHAPTER VI. ARTIFICIAL INTELLIGENCE-POWERED BRANDING IN THE METAVERSE AGE: VIRTUAL REALITY AND NEXTGENERATION CONSUMER EXPERIENCES 85 Kemal Gökhan NALBANT & Sevgi AYDIN CHAPTER VII. RECENT GLOBAL FINANCIAL CONDITION PROFILES: GLOBAL FISCAL POLICY PRIORITIES AND UNCERTAINTIES TO THE POLICIES 95 Ahmet Niyazi ÖZKER CHAPTER VIII. PREDICTIVE EMPLOYEE RETENTION ANALYTICS: DEVELOPING DATA-DRIVEN TALENT MANAGEMENT STRATEGIES THROUGH SURVIVAL ANALYSIS 115 Ayşe BOSTAN & Emel DOĞAN & Yavuz Selim BALCIOĞLU CHAPTER IX. DEVELOPMENT OF PUBLIC BANKS IN THE TURKISH BANKING SYSTEM BETWEEN 2005 AND 2024 141 MİNİRE KIRBAŞLI THE IMPACT OF THE 1925 FRIENDSHIP AND NEUTRALITY . . .   7 It is clear from İnönü›s statements that the Treaty was not a result of the USSR’s specific policy towards Turkey, but rather a necessity of the general foreign policy strategy being implemented. Furthermore, it was not only in line with the USSR’s interests, but also with Turkey’s foreign policy strategy. Foreign Minister Tevfik Rüştü Aras also assessed the 1925 Treaty, when it was ratified by the Grand National Assembly of Turkey, as a document expressing solidarity between two countries fighting against expansionist and invasionist policies (Bilge 1992). 4. The 1925 Turkey-USSR Friendship and Neutrality Treaty and the Parties’ Assessments In 1924, the USSR Ambassador to Ankara, Surizt, proposed to Prime Minister İsmet Pasha that a new treaty be drawn up, building on the 1921 Treaty, with the aim of taking relations between the two countries to a higher level. The noteworthy point here is that the proposal for the treaty came from the USSR. Traditionally, when such an agreement was under discussion, Russian foreign policy either wanted the proposal to come from the other side or presented its own proposals as if they had come from the other side (Gürün 2010). Signed in Paris on 17 December 1925 by Foreign Minister Tevfik Rüştü Aras on behalf of Turkey and Foreign Commissar Georgi Chicherin on behalf of the USSR, the Agreement consists of three articles, three protocols and a confidential letter (Soysal 1983). Article: 1. “When a third State or several States take military action against one of the Contracting Parties, the other Contracting Party undertakes to maintain its neutrality towards the former. Note: Since military manoeuvres do not harm the other Party, they should not be included within the scope of the term ‘military action’.” Article: 2. “Each Contracting Party undertakes to refrain from any form of attack against the other. Each Contracting Party undertakes not to participate in any alliance or political agreement directed against the other Contracting Party by one or more third States, or in any alliance or agreement directed against the security of the other Contracting Party on land or at sea by one or more third States. Furthermore, each Contracting Party undertakes not to participate in any hostile action directed against the other Contracting Party by one or more third States. Article 3. “This Agreement shall enter into force upon ratification and shall remain in force for a period of three years.” 8   RESEARCH, METHODS AND ANALYSIS IN SOCIAL SCIENCES . . . If the Parties do not notify their intention to terminate the Agreement six months prior to the expiry of the three-year period, the Agreement shall be deemed to be automatically extended for one year. Protocol: 1. “It is also agreed that each Contracting Party shall have complete freedom of action in its relations with third States, beyond the obligations specified in this Agreement.” Protocol: 2. “The Contracting Parties have agreed that all monetary and economic agreements between States that may be directed against the other Contracting Party shall be included within the scope of the term ‘political nature’ as set out in Article 2 of the present Agreement.” Protocol: 3. “The Contracting Parties undertake to enter into negotiations to determine and establish the form of resolution for any disputes that may arise between them and cannot be resolved through normal diplomatic channels.” In addition to the agreement and protocols, Çiçerin gave Tevfik Rüştü Aras a confidential letter.1 “Dear Minister, I wish to state that the Agreement signed today between the two Governments is complementary in nature and that the sincere friendship that has existed between the two sides since the Agreement signed in Moscow on 16 March 1921 will not be disrupted and will form the basis of their relations in the event that one of the Contracting Parties enters into war with a third State or several States.” The Treaty entered into force on 26 June 1926. The Protocol on the Extension of the Term of the 1925 Treaty was signed on 17 December 1929 during the visit of Karahan, Deputy People’s Commissar for Foreign Affairs of the USSR, to Ankara. The protocol, consisting of three articles and a secret additional protocol, entered into force on 28 July 1930. The articles of the protocol (Soysal 1983): Article: 1. “The Treaty of Friendship and Neutrality between the Republic of Turkey and the Union of Soviet Socialist Republics, signed in Paris on 17 December 1925, has been extended for a period of two years from the date of its expiry. However, if one of the Contracting Parties does not notify the other Party of its intention to terminate the Treaty six months 1 The first instance of such a letter being issued was revealed in the study entitled “Montreux and the Pre-War Years (1935-1939),” published by the Ministry of Foreign Affairs on the occasion of the 50th anniversary of the Republic (Soysal 1983). THE IMPACT OF THE 1925 FRIENDSHIP AND NEUTRALITY . . .   9 before the end of the two-year period, the Treaty shall be deemed to have been extended for a further period of one year.” Article 2. “Each Party declares that it has no obligations, other than those published in treaties, towards other States that are direct neighbours of the other Party by land or sea. Each Party undertakes not to enter into negotiations aimed at concluding political treaties with States directly neighbouring the other Party by land or sea without notifying the other Party, and to conclude such treaties only with the consent of the Party concerned. Of course, treaties to be published with the aim of establishing or maintaining normal relations with these States shall be excluded from the above obligation. Article 3. “This Protocol, which shall be annexed to the Turkish-Soviet Friendship and Neutrality Agreement and form part thereof, shall remain in force for the duration of the Agreement as extended in accordance with the provisions of Article 1 above. This Protocol shall enter into force upon the notification of ratification by the Contracting Parties.” Additional Protocol Confidential Clause: “With regard to the Protocol signed today, the undersigned agree that the following are the direct neighbours of the two States by land and sea. The neighbours of the Republic of Turkey: Iran, Iraq, Bulgaria, Greece, Italy, the British Empire, and the authority acting on behalf of Syria (meaning Mandatory France); Neighbouring States of the Union of Soviet Socialist Republics: Iran, Afghanistan, China, Mongolia, Japan, Finland, Estonia, Latvia, Poland, Romania and the British Empire.” Technically speaking, the 1925 Treaty went beyond being a document regulating relations between the two countries. It introduced a system of bilateral agreements for neutrality and non-aggression in order to ensure peace on the international stage (Vandov, 2014). When considered within the framework of a sovereign state’s right to pursue an independent foreign policy, the second article of the 1929 extension protocol, even though the principle of reciprocity applied, was a provision that restricted the ability of Turkish foreign policy to make independent decisions. This is because Turkey acted in accordance with this article in its relations with third parties. For example, in 1932, it consulted with the USSR on its membership of the League of Nations and made a reservation regarding sanctions against the USSR. The 1934 Balkan Pact was concluded in consultation with the USSR, 10   RESEARCH, METHODS AND ANALYSIS IN SOCIAL SCIENCES . . . and again a reservation was made in favour of the USSR. and the Soviet Union was consulted during the preparatory process for the 1937 Sadabat Pact and the 1939 tripartite alliance with Britain and France. However, the Soviet Union did not pursue a foreign policy consistent with the provisions of the 1925 Treaty and its additional protocols. For example, the USSR did not consult Turkey when it signed the Mutual Assistance Agreement with France in 1935 or during the alliance negotiations with Bulgaria in 1940 (Soysal, 1983). The USSR’s stance once again confirmed the reality that there should be no asymmetry of power between states involved in such relations in international politics. On 24 December 1925, Litvinov, Deputy People’s Commissar for Foreign Affairs of the USSR, explained the reasons for the 1925 Agreement in a statement to the Pravda newspaper as follows: “The 1925 Agreement has strengthened the friendly relations between the two countries and, at the same time, contributed to the maintenance of peace in the international arena in general. The agreement was not directed against any state, nor did it threaten anyone’s interests” (Degras, 1952). 5. The Impact of the 1925 Agreement on Turkey-USSR Relations The impact of the 1925 Treaty on relations between the two countries was generally positive. Bilge (1992) states that the Treaty had the effect of increasing consultation and solidarity between the two countries and that, as its duration was extended unless notice of termination was given, it remained in force until it was terminated by the USSR in 1945, which was significant for Turkey’s independence. Apart from this assessment, the first impact of the 1925 Treaty on relations between the two countries was the opening of the new USSR Embassy building in Ankara on 19 April 1926 (Vandov, 2014). Speaking at the opening ceremony, the USSR Ambassador to Ankara, Suriç, and Foreign Minister Aras emphasised in their statements that, in accordance with the spirit of the 1925 Treaty, both countries wished to develop their relations on a reciprocal basis (Vandov, 2014). In this context, as stated in the 1921 and 1925 Treaties, the Turkey-USSR General Protocol on the determination of the borders between the two countries was signed on 31 May 1926 (Vandov, 2014). Subsequently, in June 1927, an agreement on the joint use of the border and a protocol on the construction of a dam on the Aras River were signed (Vandov, 2014). Maksim Litvinov, Deputy People’s Commissar for Foreign Affairs of the Soviet Union, made important assessments regarding the 1925 Treaty THE IMPACT OF THE 1925 FRIENDSHIP AND NEUTRALITY . . .   11 in a statement published in the Pravda newspaper on 24 December 1925 (Degras, 1952). “The 1925 Treaty strengthens the long-standing de facto relations between the two countries. These relations have made it highly unlikely that the Soviet Union or Turkey would harbour any aggressive or hostile intentions towards each other. I will not deny that I regard the formal consolidation of these relations and the acceleration of the signing of the treaty as a partial response to the unfounded allegations that have appeared from time to time, and particularly frequently in the Anglo-American press recently, that the Soviet Union and Italy have concluded an agreement against Turkey. On the other hand, the Soviet Government was aware of the efforts of certain states to draw Turkey into an alliance hostile to the Soviet Union. The signing of the treaty should put an end to such alarming rhetoric and dispel any fears or doubts among the peoples of both countries about the steadfastness of Soviet-Turkish friendship. While strengthening relations between the USSR and Turkey, this Treaty also represents an important step towards securing peace. As can be understood from the text, this Treaty is not directed against anyone and does not threaten anyone’s interests. I believe it is necessary to clearly state that there are no secret clauses or protocols. The best proof of the Agreement’s peaceful intentions is that the Soviet Government is willing to conclude similar agreements with all countries with which it maintains normal diplomatic relations. The only way to eliminate the possibility of hostile groups and combinations forming is for all states to conclude agreements similar to the Turkish-Soviet Agreement. The impact of the 1925 Treaty on relations between the two countries can be assessed in two dimensions: political and economic. Politically, the Treaty had positive effects on the relations of both countries with third parties and with each other. After the Treaty, both countries developed their effectiveness in the international arena and entered a process of normalisation by resolving their issues with Western countries. After the 1925 Treaty, a process that could be called a “period of balanced relations” began between the two countries (Bilge, 1992). This process continued until the 1936 Montreux Straits Conference. The reason this period, spanning approximately ten years, is defined as a period of balanced relations is primarily because tensions between the two countries had relatively decreased and a certain balance and stability had been achieved (Bilge, 1992). 12   RESEARCH, METHODS AND ANALYSIS IN SOCIAL SCIENCES . . . In the context of the impact of the 1925 Treaty on relations between the two countries, the most significant step was the Odessa meeting between Foreign Ministers Çiçerin and Aras in 1926. The Odessa talks, which were of great importance for the development of relations between the two countries after the 1925 Treaty, focused on three main issues. The signing of a trade agreement between the parties, membership of the League of Nations, and Italy’s threat to Turkey (Tuncer, 2017). Parallel to the warming political relations between the two countries, rapprochement continued in the economic sphere. In this context, a trade agreement was concluded between Turkey and the USSR in 1927. 6. Conclusion Why was the 1925 Agreement necessary when the 1921 Moscow Agreement already existed? The answer that can be given within the framework of the findings we obtained regarding the primary problem of the study is that the concept of non-aggression was also included in the framework of friendship and brotherhood relations outlined between the parties in the 1921 Agreement. Furthermore, the changing conditions in international politics between 1921 and 1925 and the parties’ efforts to adapt their foreign policies to these changes made the 1925 Treaty necessary. In this context, the Locarno Treaties and the reaction to them, along with the USSR’s strategy of non-aggression and neutrality with neighbouring countries, constituted the reasons for such a treaty from the USSR’s perspective. From Turkey’s perspective, dissatisfaction with the League of Nations’ decision on Mosul and the strategy of pursuing a balanced foreign policy necessitated the 1925 Treaty. Therefore, the 1925 Treaty emerged as a natural consequence of the foreign policies that both countries sought to implement in the 1920s. The impact of the 1925 Treaty on Turkey-USSR relations, in relation to the second issue of the study, shows that, in general terms, the Treaty added a new dimension to TurkeyUSSR relations. The Treaty produced positive results for both countries. In terms of the foreign policy of the Republican era, Turkey had made establishing good relations with the USSR, without distancing itself from the West, one of its fundamental objectives. This goal was achieved during the period of balanced relations established with the USSR after the 1925 Treaty, which lasted for approximately ten years. However, from Turkey’s perspective, the 1925 Treaty with the USSR did not signify a radical break with the West. Turkey did not stray from the principle of Westernisation, one of the THE IMPACT OF THE 1925 FRIENDSHIP AND NEUTRALITY . . .   13 two fundamental principles of Turkish foreign policy alongside the status quo. The conditions of the international system in the first half of the 1920s, as well as the foreign policy and security interests of both countries, necessitated such a treaty between the parties. In fact, the 1925 Treaty, which was made against the West, paradoxically brought both countries closer to the West. Historically, the rhetoric that “whenever there is a Turkish-Russian rapprochement, the West will also be close to this rapprochement” has once again been applied in a manner consistent with realpolitik. References Armaoğlu, F. (1999). 20. yüzyıl siyasi tarihi, 11th Edition, Alkım Publishing. Ateş, N. Y. (2008). Cumhuriyet dönemi Türk dış politikaları, M. Bıyıklı (Ed.), Türk dış politikası: cumhuriyet dönemi in (p. 53-90), Bilimevi Press. Aydın, M. (1999). Determinants of Turkish foreign policy: historical framework and traditional inputs, Middle Eastern Studies 35 (4): 152-186. Bilge, S. (1992). Güç komşuluk, Türkiye İş Bankası Cultural Publications. Carr, E. H. (1965). International relations between the two world wars (1919-1939), Macmillan Press. Degras, J. (1952). Soviet documents on foreign policy, Vol: 2, Oxford University Press. Gönlübol, M. ve Sar. C. 1919-1939 yılları arasında Türk dış politikası, M. Gönlübol (Ed.), Olaylarla Türk dış politikası (1919-1995) in (p. 3-133), Siyasal Publishing Gürün, K. (2010). Türk-Sovyet İlişkileri (1920-1953), Türk Tarih Kurumu Publishing. Hale, W. (2003). Türk dış politikası 1774-2000, (Trans: Petek Demir), Mozaik Publishing. İnönü, İ. (1987). Hatıralar, Vol: 2, Bilgi Publishing. Kissenger, H. (2015). Diplomacy, 14th Edition, (Trans: İbrahim H. Kurt), Türkiye İş Bankası Cultural Publications. Küçük, E. ve Bekar, N. (2019). Türk diplomatların bakış açısıyla TürkSovyet ilişkileri (1927-1925). Ankara Üniversitesi SBF Dergisi, 74 (1), 163192. Pick, O. (1967). Soviet Alliance Policies in Retrospect. International Journal, 22 (4), 576-592. 14   RESEARCH, METHODS AND ANALYSIS IN SOCIAL SCIENCES . . . https://www.jstor.org/stable/pdf/40200199. pdf?refreqid=fastlydefault%3A3ea72be02242a0e6fd57256cad351a5c&ab_ segments=&initiator=&acceptTC=1 Sander, O. (1996). Siyasi tarih (1918-1994), 5th Edition, İmge Publishing. Soysal, İ. (1983). Türkiye’nin siyasal antlaşmaları, 1. Cilt (1920-1945), Türk Tarih Kurumu Publishing. Sosyalizm ve Toplumsal Mücadeleler Ansiklopedisi, (1988), Vol: 3. İletişim Publishing. Tellal, E. (2009). SSCB’yle İlişkiler. B. Oran (Ed.), Türk dış politikası: kurtuluş savaşından bugüne olgular, belgeler, yorumlar, Vol: 1, İletişim Publishing. Tuncer, H. (2017). Türk dış politikası cumhuriyet dönemi (1920-2002), Kaynak Publishing. Vandov, D. (2014). Atatürk dönemi Türk-Sovyet ilişkileri, Kaynak Publishing. Yüceeer, S. (2011). Atatürk dönemi (1919-1938) Türk-Rus ilişkilerinin siyasi boyutu, İ. Kamalov ve İ. Svistunova (Ed.), Atatürk’ten soğuk savaş dönemine Türk-Rus ilişkileri in (p. 61-106), Atatürk Araştırma Merkezi. 15 CHAPTER II AN ANALYSIS OF THE 44-DAY KARABAKH VICTORY IN THE CONTEXT OF “TWO STATES ONE NATION” IN TÜRKİYE– AZERBAIJAN RELATIONS Elnur Hasan MİKAİL1 & Hakan ÇORA2 1 (Prof. Dr.), Kafkas University, Kars E-mail: [email protected]g 0000-0001-9574-4704 2 (Assoc. Prof. Dr.), Maltepe University, Istanbul E-mail: [email protected] 0000-0001-5780-549X 1. Introduction The relationship between Türkiye and Azerbaijan occupies a unique position in international relations because it is not solely based on pragmatic geopolitical or economic interests, but also on a profound sense of shared identity. The popular phrase “İki devlet, bir millet” (“Two States, One Nation”) is more than a slogan: it encapsulates a worldview that considers the destinies of Türkiye and Azerbaijan as intrinsically intertwined (Murinson, 2010). While kinship, linguistic affinity, and cultural heritage have always underpinned this sentiment, the political articulation of this idea has gained new significance in the twenty-first century. The 44-day war in Nagorno-Karabakh in September–November 2020 brought this principle into sharp relief. Azerbaijan’s decisive victory over Armenian forces, achieved with significant political, military, and diplomatic support from Türkiye, represented not only the restoration of Azerbaijani sovereignty over long-occupied territories but also the consolidation of a new regional security order in the South Caucasus (Cornell, 2017; Özkan, 2021). The outcome symbolized a moment in which the rhetoric of “Two States, 16   RESEARCH, METHODS AND ANALYSIS IN SOCIAL SCIENCES . . . One Nation” became materially manifest on the battlefield and in post-war diplomacy. The purpose of this chapter is to analyze the war and its consequences for Turkish–Azerbaijani relations. It proceeds on several levels. First, it contextualizes the modern alliance within the historical relationship between Anatolian Turks and Azerbaijanis, tracing continuities from the Ottoman and Safavid encounters through the Soviet period and into the post-1991 independence era. Second, it situates the Nagorno-Karabakh conflict in its broader historical evolution, showing how the failure of diplomacy over nearly three decades created the conditions for war. Third, it explores the conduct of the 2020 war itself, emphasizing Turkish contributions in terms of military technology, training, and political support. Finally, it assesses the diplomatic and geopolitical aftermath, arguing that the war consolidated Turkish–Azerbaijani cooperation into a de facto alliance with long-term implications for Eurasian geopolitics (Shaffer, 2021). The chapter contributes to the academic debate on the transformation of regional security orders by demonstrating how identity-driven alliances can shape outcomes of military conflicts and restructure post-war diplomacy. Unlike alliances formed primarily on transactional bases, the Turkish–Azerbaijani partnership is deeply rooted in identity, culture, and shared narratives of struggle and victory. The 44-day war confirmed that such partnerships can have decisive outcomes when combined with technological innovation and coherent strategic planning. 2. Historical Background 2.1 Ottoman–Caucasian Connections The historical links between Anatolia and the Caucasus predate the formation of modern states. The Ottoman Empire, as the preeminent TurkicMuslim power in the region, had periodic confrontations and accommodations with Persian and Russian forces over the Caucasus. The Safavid–Ottoman rivalry in the sixteenth and seventeenth centuries frequently involved control over the territories of present-day Azerbaijan. The 1639 Treaty of Zuhab, for example, demarcated the spheres of influence between the Ottomans and the Safavids, with much of modern Azerbaijan falling under Persian control (Cornell, 2017). Despite political boundaries, the peoples of Anatolia and Azerbaijan retained close linguistic, cultural, and religious ties, facilitated by trade routes, migration, and intermarriage. AN ANALYSIS OF THE 44-DAY KARABAKH VICTORY IN THE . . .   23 Prime Minister Nikol Pashinyan’s 2019 declaration that “Karabakh is Armenia, period” eliminated any remaining ambiguity in Yerevan’s position and was perceived in Baku as a categorical rejection of compromise (De Waal, 2013). The July 2020 clashes in Tovuz further hardened Azerbaijani resolve. Public protests in Baku, where tens of thousands chanted “Karabakh or death,” illustrated that popular patience had evaporated (Cornell, 2017). At the same time, Azerbaijan’s substantial military modernization, including the acquisition of Turkish and Israeli drones, shifted the balance of power decisively in its favor (Huseynov, 2022). Thus, when clashes erupted on 27 September, Azerbaijan launched a coordinated offensive, arguing that it was exercising its right to self-defense under Article 51 of the UN Charter. Armenia declared martial law, and both sides mobilized significant forces. Unlike previous flare-ups, this escalation rapidly expanded into a full-scale war (Shaffer, 2021). 4.2 Battlefield Dynamics and Military Strategy The battlefield dynamics of the 2020 war contrasted sharply with those of the 1990s. Azerbaijan implemented a coherent strategy emphasizing mobility, precision strikes, and technological superiority, while Armenia relied heavily on entrenched defensive positions and outdated Soviet-era doctrine. Southern Front Offensive: Azerbaijan’s main thrust came in the southern sector, particularly the Jabrayil–Fuzuli axis. After initial breakthroughs, Azerbaijani forces captured key towns, gradually advancing northward. The strategy sought to avoid the heavily fortified northern approaches to Nagorno-Karabakh and instead exploit Armenia’s weaker defenses in the south (Huseynov, 2022). Encirclement of Shusha: The climax came with the liberation of Shusha on 8 November 2020. Shusha, a cultural and strategic stronghold overlooking Stepanakert (Khankendi), was considered the “heart of Karabakh.” Its capture broke Armenian morale and forced Yerevan to accept the Russia-brokered ceasefire on 9 November (De Waal, 2013). Attrition of Armenian Forces: Armenian forces suffered heavy losses in armor, artillery, and personnel. According to Azerbaijani claims corroborated by open-source intelligence, hundreds of tanks and air defense systems were destroyed by drone strikes (Gafarov & Erdoğan, 2021). In short, Azerbaijan’s battlefield success stemmed from a combination of careful planning, superior technology, and political will. 24   RESEARCH, METHODS AND ANALYSIS IN SOCIAL SCIENCES . . . 4.3 Turkish Support: Military, Political, and Diplomatic Türkiye’s support for Azerbaijan during the war was decisive and multidimensional. While Turkish troops did not directly participate in combat, Ankara’s assistance was pivotal in shaping the outcome. Military Training and Doctrine: Over the past two decades, Türkiye had trained thousands of Azerbaijani officers and conducted joint exercises. This exposure to NATO-style operational planning significantly improved Azerbaijani command structures (Murinson, 2010). Weapons and Technology: Turkish-made Bayraktar TB2 drones became the symbol of the war. They enabled Azerbaijan to systematically degrade Armenian defenses by targeting tanks, artillery, and supply convoys (Huseynov, 2022). Political Backing: President Recep Tayyip Erdoğan’s unequivocal support emboldened Baku. Ankara framed the conflict as a matter of justice and international law, consistently emphasizing Azerbaijan’s right to reclaim its occupied lands (Özkan, 2021). Diplomatic Maneuvering: Türkiye’s presence at the post-war monitoring center in Aghdam signaled a permanent role in the regional security architecture, breaking Russia’s monopoly as the sole mediator (Shaffer, 2021). The war also strengthened Türkiye’s global image as a rising defense exporter. International observers noted that the “drone warfare model” pioneered in Karabakh would influence conflicts from Libya to Ukraine (Huseynov, 2022). 4.4 The Role of Technology and Innovation The 44-day war has been described as the world’s first “drone-dominated conflict.” Azerbaijan’s integration of UAVs, electronic warfare, and satellite intelligence transformed the battlefield. Unmanned Aerial Vehicles (UAVs): Bayraktar TB2s and Israeli-made Harop loitering munitions systematically destroyed Armenian tanks, air defenses, and artillery. This “aerial hunting campaign” neutralized Armenia’s capacity to maneuver (Huseynov, 2022). Electronic Warfare: Azerbaijan employed advanced electronic jamming to suppress Armenian communications, further amplifying the effectiveness of UAV strikes (Cornell, 2017). Open-Source Intelligence (OSINT): Azerbaijan strategically released drone footage, shaping both domestic morale and international perception. The AN ANALYSIS OF THE 44-DAY KARABAKH VICTORY IN THE . . .   25 dissemination of videos depicting Armenian losses became an effective tool of psychological warfare (Shaffer, 2021). Combined Arms Integration: Unlike in the 1990s, Azerbaijan effectively combined artillery, special forces, and UAVs, creating a multi-layered offensive strategy. This technological dominance rendered Armenian Soviet-era equipment largely obsolete. Analysts noted that the war demonstrated how a smaller state, with the right technology and doctrine, could defeat entrenched defenses (Gafarov & Erdoğan, 2021). 4.5 Humanitarian Dimensions of the War The war also had a profound humanitarian impact. Official estimates suggest more than 6,500 soldiers were killed on both sides, with civilian casualties numbering in the hundreds (De Waal, 2013). Over 100,000 Armenians fled Nagorno-Karabakh during hostilities, creating new displacement crises. For Azerbaijan, the return of IDPs to liberated territories became both a logistical challenge and a national priority (Cornell, 2017). International organizations, including the UN and ICRC, struggled to respond due to limited access. Allegations of war crimes, such as indiscriminate shelling of civilian areas in Ganja and Stepanakert, underscored the brutality of modern warfare. Nevertheless, the war ended with Azerbaijan regaining control of most occupied districts and establishing the conditions for large-scale reconstruction projects (Shaffer, 2021). 4.6 The Ceasefire Agreement of 10 November 2020 On 9–10 November, after the fall of Shusha, Armenia agreed to a ceasefire brokered by Russia. The agreement included: Withdrawal of Armenian forces from occupied districts. Deployment of 2,000 Russian peacekeepers to Nagorno-Karabakh. Establishment of a transport corridor linking Azerbaijan proper to Nakhchivan via Armenia’s Syunik region. Creation of a joint Russian–Turkish monitoring center (Özkan, 2021). For Azerbaijan, this was a decisive victory that restored sovereignty over significant territories. For Armenia, it was a devastating defeat that triggered a political crisis in Yerevan. For Türkiye, it was an affirmation of its rising role as a regional security guarantor alongside Russia. 26   RESEARCH, METHODS AND ANALYSIS IN SOCIAL SCIENCES . . . 5. Diplomatic and Geopolitical Implications The end of the 44-Day War not only redrew the map of the South Caucasus but also reshaped the region’s diplomatic order. The ceasefire, though brokered by Russia, was profoundly shaped by Turkish influence. The war revealed the limits of Western diplomacy, weakened the OSCE Minsk Group, and demonstrated that hard power combined with identity-driven alliances could create new political realities. 5.1 Consolidation of the Turkish–Azerbaijani Alliance One of the most significant consequences of the war was the consolidation of the Turkish–Azerbaijani strategic partnership into what increasingly resembles an alliance. The Shusha Declaration of June 2021 formally enshrined this transformation. Signed by Presidents Ilham Aliyev and Recep Tayyip Erdoğan in the historic city of Shusha, the declaration committed the two states to mutual defense cooperation, economic integration, and cultural solidarity (Aliyev, 2021). For Azerbaijan, Türkiye’s role in the war provided concrete evidence that Ankara would stand by its side in times of crisis. For Türkiye, Azerbaijan’s victory validated decades of investment in bilateral ties and strengthened its credibility as a regional power (Murinson, 2010). The declaration symbolized the institutionalization of the “Two States, One Nation” principle into binding commitments. This consolidation also extended into military affairs. Joint exercises became more frequent, and interoperability between Turkish and Azerbaijani forces deepened. Ankara emerged as Baku’s main defense partner, gradually replacing Moscow’s role as Azerbaijan’s primary supplier and trainer (Gafarov & Erdoğan, 2021). 5.2 Russia’s Strategic Calculations Russia, long the dominant actor in the South Caucasus, faced a more complex environment after the war. On one hand, Moscow succeeded in brokering the ceasefire and deploying peacekeepers to Nagorno-Karabakh, thereby preserving its leverage over both sides. On the other hand, Türkiye’s entry into the post-war order signaled that Russia could no longer monopolize regional security (Özkan, 2021). Russian policymakers had historically sought to maintain a balance between Armenia and Azerbaijan, ensuring neither side gained full victory. The 2020 war disrupted this balance, as Azerbaijan’s success reduced Yerevan’s AN ANALYSIS OF THE 44-DAY KARABAKH VICTORY IN THE . . .   27 reliance on Moscow while increasing Baku’s confidence. The presence of Turkish personnel at the joint monitoring center further eroded Russia’s exclusivity (Shaffer, 2021). For Moscow, this was a double-edged sword. While it preserved shortterm influence, it also faced the prospect of long-term decline as Ankara’s role expanded. The Russian invasion of Ukraine in 2022 further distracted Moscow, weakening its ability to sustain dominance in the Caucasus (Cornell, 2017). 5.3 Iran’s Uneasy Position Iran reacted uneasily to Azerbaijan’s victory and Türkiye’s enhanced role. Tehran feared that the opening of the Zangezur Corridor might weaken its own position as a transit hub and strengthen pan-Turkic connectivity (Shaffer, 2021). Moreover, Iran has a sizable Azerbaijani minority, estimated between 15–20 million people, whose ethno-linguistic ties to Azerbaijan could inspire cultural assertiveness. During the war, Iranian officials attempted to balance neutrality with tacit support for Armenia. Post-war, Tehran intensified military exercises along its northern border, signaling discomfort with Turkish–Azerbaijani cooperation. Nevertheless, Iran was compelled to acknowledge the new reality and adjust its policies to prevent further marginalization (Özkan, 2021). 5.4 Western Powers and the Decline of the Minsk Group The war exposed the irrelevance of the OSCE Minsk Group. Despite nearly three decades of mediation, the co-chairs — the United States, France, and Russia — had failed to deliver tangible progress. In 2020, Washington was distracted by domestic political turmoil, Paris was perceived in Baku as biased toward Armenia, and Russia prioritized its role as security broker (De Waal, 2013). The result was the marginalization of Western influence in the South Caucasus. Türkiye and Russia emerged as the primary power brokers, while the EU and U.S. became reactive rather than proactive. This shift demonstrated that in unresolved conflicts, military outcomes can supplant diplomatic stagnation (Shaffer, 2021). 6. “Two States, One Nation” in Practice The war and its aftermath elevated the “Two States, One Nation” principle from rhetorical slogan to political reality. This transformation can be observed in multiple domains. 28   RESEARCH, METHODS AND ANALYSIS IN SOCIAL SCIENCES . . . 6.1 Cultural and Societal Solidarity The Azerbaijani victory was celebrated in both Baku and Ankara as a shared triumph. Turkish flags waved alongside Azerbaijani ones during victory parades, symbolizing unity at the societal level (Cornell, 2017). Public opinion polls in both countries showed overwhelming support for closer integration. Cultural exchanges deepened. Joint academic conferences, media productions, and artistic collaborations framed the war as a common story of resilience. Narratives of martyrdom and sacrifice were invoked in both states, reinforcing the sense of kinship (Shaffer, 2021). 6.2 Economic and Energy Cooperation The war reinforced the strategic value of existing energy corridors and motivated new projects. The Trans-Anatolian Natural Gas Pipeline (TANAP) and Trans-Adriatic Pipeline (TAP) already connected Azerbaijani resources to European markets through Türkiye (Ismayilov, 2021). Post-war, these projects gained greater security as Azerbaijan’s territorial sovereignty was restored. The reconstruction of liberated territories also created opportunities for Turkish firms, which won contracts in infrastructure, housing, and transportation. This economic partnership underscored how military victory translated into long-term economic integration (Gafarov & Erdoğan, 2021). 6.3 Security and Defense Integration Perhaps the most visible manifestation of “Two States, One Nation” has been in defense. The Shusha Declaration institutionalized joint military exercises, arms production, and intelligence cooperation (Aliyev, 2021). Turkish defense companies such as Baykar and ASELSAN expanded their partnerships in Azerbaijan, fostering indigenous capacity. Furthermore, the two countries began to coordinate security policy beyond the bilateral level, extending cooperation to multilateral Turkic frameworks. The Organization of Turkic States increasingly discussed collective defense and regional security, inspired by the Azerbaijani–Turkish example (Shaffer, 2021). 6.4 Regional Turkic Solidarity The war galvanized Turkic solidarity across Central Asia. Leaders in Kazakhstan, Uzbekistan, and Kyrgyzstan openly expressed support for Azerbaijan during the conflict. The Organization of Turkic States, rebranded in AN ANALYSIS OF THE 44-DAY KARABAKH VICTORY IN THE . . .   29 2021, took on new significance as a platform for political, cultural, and security cooperation (Cornell, 2017). For Ankara, this was part of a broader strategic vision: positioning Türkiye as the anchor of the Turkic world, with Azerbaijan as its critical partner and gateway to Central Asia. For Baku, Turkish solidarity enhanced its strategic depth and regional influence (Özkan, 2021). 7. Post-War Regional Order The 44-Day War not only transformed the relationship between Türkiye and Azerbaijan but also redefined the strategic architecture of the South Caucasus. The ceasefire of November 2020 introduced new geopolitical dynamics that continue to unfold. 7.1 The Zangezur Corridor Debate One of the most contested provisions of the ceasefire was the reopening of transportation links, including a potential corridor through Armenia’s Syunik region, often referred to as the Zangezur Corridor. For Azerbaijan, this route would provide direct land access to its exclave of Nakhchivan and, through it, to Türkiye. Such a corridor would complete a strategic East–West axis, linking the Caspian to Anatolia without reliance on Iran (Ismayilov, 2021). For Türkiye, the Zangezur Corridor is more than logistics: it represents the physical realization of pan-Turkic connectivity, enabling direct routes to Central Asia. For Armenia, however, the corridor raises sovereignty concerns, as it fears being bypassed or reduced to a transit state. Iran also opposes the project, worried that its own transit role will diminish (Özkan, 2021). If implemented, the Zangezur Corridor could revolutionize Eurasian trade and energy flows, embedding Türkiye and Azerbaijan as central nodes in the Silk Road revival. However, as of 2023, political disagreements and security concerns have delayed its full realization (Shaffer, 2021). 7.2 A New Security Architecture in the South Caucasus The war altered the balance of power in the South Caucasus in several ways: Russia’s Waning Monopoly: The deployment of peacekeepers preserved Moscow’s leverage, but Türkiye’s permanent presence marked the end of Russia’s exclusive role (Cornell, 2017). 30   RESEARCH, METHODS AND ANALYSIS IN SOCIAL SCIENCES . . . Türkiye as a Co-Guarantor: Ankara’s role in post-war arrangements elevated its profile as a legitimate security actor in the Caucasus (Gafarov & Erdoğan, 2021). Azerbaijan’s Ascendancy: The victory cemented Azerbaijan as the region’s most powerful state, able to shape outcomes and project influence beyond its borders (Shaffer, 2021). This emerging architecture is multipolar rather than unipolar, with Türkiye and Russia as co-brokers, Iran as a wary observer, and Western powers largely sidelined. 7.3 Azerbaijan as a Gateway to Central Asia Post-war, Azerbaijan positioned itself as a geopolitical bridge between Türkiye and Central Asia. The Middle Corridor (Trans-Caspian International Transport Route) gained momentum as an alternative to Russian-dominated northern routes and Chinese-controlled southern routes (Ismayilov, 2021). Türkiye and Azerbaijan jointly promoted this corridor, linking Chinese and Central Asian markets through the Caspian Sea, Azerbaijan, Georgia, and Türkiye to Europe. In this vision, Azerbaijan becomes the linchpin of East–West connectivity, enhancing Ankara’s strategic outreach to the Turkic world (Özkan, 2021). 8. Conclusion The 44-Day Karabakh War was a transformative event in the history of Türkiye–Azerbaijan relations. It represented the culmination of decades of shared cultural solidarity, political cooperation, and strategic alignment. By enabling Azerbaijan to restore its territorial integrity, the war validated the “Two States, One Nation” principle not just as rhetoric but as a functional framework for alliance politics. Türkiye’s role was decisive: politically, Ankara’s unwavering support gave Azerbaijan confidence; militarily, Turkish technology and training shaped the battlefield; diplomatically, Türkiye secured a permanent place in the post-war order. The Shusha Declaration institutionalized this partnership, elevating it into a formal alliance. The war also restructured the South Caucasus. Russia remains influential but no longer hegemonic; Iran is uneasy and defensive; Western actors have receded. Meanwhile, Türkiye and Azerbaijan stand at the center of a new AN ANALYSIS OF THE 44-DAY KARABAKH VICTORY IN THE . . .   31 regional architecture that extends toward Central Asia. The Zangezur Corridor and the Middle Corridor projects illustrate how military victory translates into long-term geoeconomic influence. Ultimately, the war demonstrated that identity-driven alliances — rooted in kinship, language, and shared destiny — can wield transformative power in international politics. The Turkish–Azerbaijani partnership, forged in history and tested in war, has emerged as a model of strategic solidarity in an era of shifting global orders. References Aliyev, I. (2021). Statement on the Shusha Declaration. Ministry of Foreign Affairs of Azerbaijan. Cornell, S. E. (2017). Azerbaijan since independence. Routledge. De Waal, T. (2013). Black garden: Armenia and Azerbaijan through peace and war. New York University Press. Gafarov, B., & Erdoğan, M. (2021). The Turkish–Azerbaijani military partnership and the Karabakh war. Journal of Eurasian Studies, 12(2), 145–162. https://doi.org/10.1177/1879366521100231 Huseynov, R. (2022). The role of drone warfare in Azerbaijan’s victory. Defense & Security Analysis, 38(1), 33–50. https://doi.org/10.1080/14751798 .2022.2021345 Ismayilov, M. (2021). Energy security and Turkish–Azerbaijani cooperation after the 2020 war. Caucasus International, 11(1), 21–39. Kucera, J. (2020). Turkey’s role in the Karabakh conflict. Eurasianet. Retrieved from https://eurasianet.org Murinson, A. (2010). The strategic importance of the Turkish–Azerbaijani partnership. Middle Eastern Studies, 46(5), 697–710. https://doi.org/10.1080/0 0263206.2010.492971 Özkan, M. (2021). The geopolitics of the 44-Day War: Türkiye, Russia, and Iran in the South Caucasus. Insight Türkiye, 23(1), 55–72. https://doi. org/10.25253/99.2021231.05 Shaffer, B. (2021). The new geopolitics of the South Caucasus after the Karabakh war. Foreign Policy Analysis, 17(4), 1–19. https://doi.org/10.1093/ fpa/orab005 Yunusov, A. (2020). Karabakh: Past, present, future. Baku Research Institute. THE MONETARY POLICY OF THE EUROPEAN CENTRAL BANK . . .   39 trustworthiness shaped the early years of one of the world’s most important monetary authorities. Surico (2007) gives a thorough factual and theoretical look at how the ECB worked in its first six years. The main idea behind the study is a nonlinear policy response function that takes into account the fact that the ECB might not want the same things when it comes to inflation, output, and keeping interest rates stable. The study goes against the usual linear-quadratic models by considering convexities in both the ECB’s loss function. The results show that the ECB has a similar response to inflation going above or below its goal, but it has different reactions to changes in output and interest rates. It reacts more strongly to drops in output than to rises in output of the same size. The study builds a version of the New Keynesian model by adding irregularities and nonlinearities to both the aggregate supply curve and the goal function of the central bank. It is thought of that the ECB’s policy is like an answer to an intertemporal optimization problem. The central bank sets interest rates to minimize a generalized loss function that has exponential fines for changes in inflation, output, and interest rates. This lets the study check if changes in economic variables that are above or below goal values are dealt with in different ways. The Generalized Method of Moments (GMM) is used to assess data from the euro-area from 1999 to 2004 in the empirical method. The results from the experiments show that the ECB’s policy actions are not symmetrical, especially when it comes to the output gap. The results show that output declines cause a bigger easing reaction than counterparty increases of the same size. Even though inflation seems to be dealt with equally, the central bank’s reaction is stronger when output is significantly higher than capacity. This shows that the economy is not structured in a straight line. The ECB also changes policy rates more quickly when the current interest rate is higher than its implicit goal. This shows that it doesn’t like interest rates that are too high. This difference in interest rates helps us understand why rates stay low and why there is more housing after a slump. Lastly, the study looks into how monetary aggregates, especially the M3 growth rate, fit into the ECB’s plan. Even though money is a big part of the ECB’s two-pillar method, the results show that M3 doesn’t play a role in policymaking on its own. On the contrary, it is used as a measure to help predict inflation. The tests that use different output gap measures show that asymmetric responses to real activity persist. This supports the idea that the ECB’s early policy framework, even though it seemed symmetric and based on rules, actually had nonlinear and state-dependent behavior. This more complex knowledge adds a lot to what is 40   RESEARCH, METHODS AND ANALYSIS IN SOCIAL SCIENCES . . . known about how to make monetary policy and how central banks should act in a newly combined currency area. Sauer and Sturm (2007) look into whether Taylor-type interest rate rules can be used to judge the policy choices made by the European Central Bank. The study’s goal is to find out if the ECB has used a stabilizing or destabilizing method to monetary policy by estimating different Taylor rules models using real-time and poll data. It says that traditional studies might get the ECB’s behavior wrong if they don’t take into account how forward-looking modern monetary policy is. This would give an incomplete picture of the ECB’s overall strategy. The results of the simultaneous rule estimates, in which interest rates are modeled based on real-time data on output gap and inflation, show that the ECB has not responded sufficiently to changes in inflation. This shows that policy is not stable because changes in nominal interest rates don’t raise real interest rates enough to counteract inflation, which goes against the Taylor principle. The Bundesbank, on the other hand, has been shown to have used more active and inflation-targeting tactics. Furthermore, the research shows that current Taylor rules often have statistical flaws like serial correlation and non-stationarity that make them less reliable when used without any modifications. To get around these problems, the authors add parts of the Taylor rule structure that look ahead. To show how the ECB makes decisions based on predictions, they use both rational expectations and real-time poll data. The results of these models are very different. If you assume forward-looking behavior, the ECB seems to follow a more calming policy rule, as shown by inflation factors above one. This is in line with the Taylor principle. The best empirical fit comes from combining poll data from the past and data collected in the present. This suggests that expectationsbased data better reflect the ECB’s strategic direction than current economic indicators. In the end, the study gives a complex assessment of ECB policy and warns against oversimplified or fixed views of its response function. When looking back at data-driven current studies, the ECB may seem passive. But when looking ahead, they show a more assertive and inflation-focused behavior. The results show how important it is to model the actions of central banks in ways that take into account current information, changes to data, and expected changes. The study shows how hard it is to evaluate monetary policy in real life and confirms that Taylor-type rules can be useful as both descriptive tools and policy measures when they are used correctly. Claeys and Demertzis (2017) give a detailed look at the European Central Bank’s plans to move away from the unusual monetary policies that were used THE MONETARY POLICY OF THE EUROPEAN CENTRAL BANK . . .   41 during and after the 2008 global financial crisis. The study initially discusses the alterations in the ECB’s operations due to prolonged low inflation, financial instability, and the necessity for interest rates to remain below zero. It stresses that going back to the way things were before the crisis might not be possible or even desirable because the euro area economy and monetary policy have changed in important ways. Instead, the authors argue that we should change our ideas about what a “normal” monetary policy is. They acknowledge that big central bank balance sheets and non-traditional tools like asset purchase plans will still be useful in the future. The study shows how the Federal Reserve’s approach of decreasing purchases of assets, raising interest rates, and then slowly shrinking the balance sheet works. Using this as an example, the authors say that the ECB should take a similar slow and clear road to keep the markets stable and the economy strong. But it also says that the euro area and the U.S. are very different in important ways, like how the banking sector is set up and how different countries’ budgets work. Because of these differences, the ECB may need to be more careful with its stabilization approach to keep things from getting worse, especially in member states that are already struggling. The authors question the idea that lowering the ECB’s balance sheet should be a top goal when they think about the best future for the bank’s monetary policy. They say that the ECB should accept a new normal in which big balance sheets are the rule, since neutral interest rates will likely stay low and quantitative easing will likely be used again in the future. Not only might it be hard to quietly get a leaner balance sheet in a reasonable amount of time, but it might also make it harder for the ECB to act to future downturns. According to the study, a central bank’s power doesn’t depend on how big its balance sheet is, but on how well it can handle cash and keep key short-term interest rates in check. The study ultimately demonstrates the significance of dialogue in the normalizing process. The ECB should not employ a rigid calendar-based approach; rather, it should adopt a policy that adapts according to inflation and GDP dynamics. The study indicates that effective communication can prevent market disruptions and enhance the ECB’s credibility. To sum up, the ECB’s efforts to normalize monetary policy will be difficult and take a long time. To be successful, the process must be flexible, well-communicated, and carefully planned, rather than a hasty return to old models. Claeys, Demertzis, and Mazza (2018) look into how the European Central Bank can change the way it makes policy as structural and macroeconomic risks rise. The first part of the study addresses significant long-term issues, such as 42   RESEARCH, METHODS AND ANALYSIS IN SOCIAL SCIENCES . . . the falling neutral interest rate, the weakening of the Phillips curve, and the fact that the euro area’s institutions aren’t fully set up yet. All of these problems make it harder for the ECB to use normal policy tools and make it harder to handle sudden changes in demand. In light of this unsure situation, the authors contend that the ECB requires a more organized and adaptable framework that can produce good results in a wide range of uncertain situations. One of the main suggestions is that the ECB should change how it defines price stability. This would make things clearer and give people more options. They say that this change would lower the chance of policy reversals happening too soon and give inflation more room to overshoot after times of undershooting. This would be especially helpful when the lower bound is close to zero. The study also says that core inflation should be targeted instead of headline inflation. It says that the ECB should be responsible for price measures it can directly affect instead of those that are affected by outside factors like energy prices. The study goes into more detail about the conflict between stable prices and stable finances, arguing that monetary policy is not the best way to directly handle financial threats. People think that interest rates are not a good way to fix problems like asset bubbles or too much loan growth. The authors advocate for increased reliance on macro prudential instruments, facilitating collaboration between national and supranational regulatory authorities, and reforming institutional operations to ensure uniform policy responses throughout the euro area. They indicate that attempting to utilize monetary policy for dual objectives concurrently may provide suboptimal outcomes, particularly if the objectives diverge. In the end, the study gives a complete plan for how the ECB can update its monetary policy tools in a world where doubt is high and institutions are scattered. It says that keeping credibility is important, but it also says that monetary government needs to change, not stay the same. This means changing strategies for targeting inflation, being open to unusual tools like negative rates or even “helicopter money,” and improving communication so that the focus isn’t on giving certainty but on showing how policy is made to handle a lot of different outcomes. As a whole, the method encourages adaptability, cooperation, and readiness as the guiding principles of an ECB structure that can handle the future. Akbakay (2018) looks at the European Central Bank’s unique plan to keep prices stable in the euro area. The author gives some historical and administrative background, focusing on how the ECB learned from past times of stagflation and inflation. When the European Central Bank (ECB) took over monetary policy in 1999, keeping prices stable was made its main goal. For THE MONETARY POLICY OF THE EUROPEAN CENTRAL BANK . . .   43 this goal to be reached, there had to be strong theory and real-world proof that long-term inflation hurts economic growth and jobs. The study goes into detail about why a medium-term perspective was chosen, focusing on how important it is for policymakers to be able to react quickly to changes in the economy and the delays that come with monetary transmission. It also shows how important a clear inflation goal is for setting standards for the public and making policy more open and accountable. The “two-pillar” analysis approach is the second most important part of the ECB’s plan. The first pillar, “economic analysis,” looks at things like wages, exchange rates, and product prices that affect price changes in the short and middle term. The second part, “monetary analysis,” focuses on how the money supply and inflation have changed over time. This two-part structure lets the ECB use a lot of different kinds of data to make policy choices. Some people have said that monetary aggregates don’t have any use in current monetary policy, but this study supports their use in better understanding the macro economy and predicting future inflation trends. This study comes to the conclusion that the ECB’s monetary policy plan is a hybrid model that has parts of both monetary targeting and inflation targeting but doesn’t fully match with either. The ECB tries to deal with both short-term changes in the economy and long-term inflationary forces by using a dual analysis method and a mediumterm outlook. The ECB’s approach is seen as a practical response to the unique challenges of a varied and complicated monetary union, even though it is still criticized. The study shows how important it is to change policy frameworks to keep prices and finances stable in a world economy that is changing quickly. Cengiz (2019) examines the strategic significance of central bank communication in enhancing the efficacy of monetary policy, using the European Central Bank as a case study. The study starts by explaining why there has been a global change toward openness in central banking since the 1990s. It is emphasized that more openness builds trust among the public and makes policy work better. The study stresses that good communication is not only necessary but also very important for policy transfer in the Euro Area, which is known for being complicated and unclear. In this way, the ECB is shown as a cuttingedge organization that communicates with different groups of people through a wide range of channels, such as news conferences, reports, talks, and interviews. The study then goes into more detail about how central bank communication has changed over time, both theoretically and practically. It shows how it has gone from being an afterthought to an important part of monetary policy. It looks at how, in the past, central banks liked keeping things secret but over 44   RESEARCH, METHODS AND ANALYSIS IN SOCIAL SCIENCES . . . time became more open as they realized how important communication is for controlling expectations. This change happened faster because of the need for more responsibility from central banks and the rise of rules for targeting inflation. The study says that good communication between central banks and the public lowers knowledge gaps, keeps financial markets stable, and grounds people’s expectations about inflation. All of these things make monetary policy more credible and predictable. The study also looks at how information affects the results of monetary policy, focusing on how it manages expectations to change shortand long-term interest rates. It shows that, clear and consistent message can be used instead of policy action to change the way the market acts. It talks about real-world evidence that backs up the idea that openness can help make markets less uncertain, especially during emergencies when other methods don’t work as well. The concept of “open mouth operations” is also examined. This is when central banks signal their intention to alter operational dynamics without implementing any actual changes. This demonstrates that communication may function independently as a policy instrument. Finally, the study looks at the ECB’s specific ways of communicating and records the range and depth of its contact efforts. It talks about how the ECB shares information through monthly news conferences, regular updates, yearly reports, speeches, working with academics, and its website. The study cites empirical studies that show the ECB’s messages do have a measurable effect on market expectations and predictions of inflation. The study ends by saying that the ECB’s model of proactive and open communication should be used as an example by other central banks, especially when it comes to handling credibility and public participation in a monetary union with people from different languages and cultures. Zabala and Prats (2020) look into the effects and ways that the ECB’s nontraditional monetary policies are transmitted, focusing on the years after the financial crisis of 2007–2008. As the zero lower bound for interest rates got closer, the ECB did things other than just setting rates. One of these was quantitative easing, which involved buying a lot of assets. The point of their study is to find out if these steps made a real difference in the Eurozone’s economic growth and price stability. The study looks at the Asset Purchase Program and uses both real and monetary factors to try to figure out what role balance sheet growth played in increasing demand and keeping inflation expectations in check. The method is based on using a Structural Vector Autoregressive (SVAR) model to look at how inflation, economic growth, and the ECB’s overall assets change over time. The model uses short-run and long-run zero limits to separate the effects of structural THE MONETARY POLICY OF THE EUROPEAN CENTRAL BANK . . .   45 shocks and find the tracks of cause and effect. The writers look at quarterly data from 2007 to 2018 and focus on the time when the ECB’s main tool was unconventional monetary policy. The model tries to figure out how well these measures work on macroeconomic indicators in the short and long term by using impulse response functions and past decompositions. The results show that the unusual policies, especially the balance sheet expansion, had a big effect on real GDP growth in the short term. This proves that they worked right after the crisis. But inflationary reactions were weaker and took longer to happen. This showed that while buying assets helped the economy rebound, they weren’t as effective at quickly stabilizing prices. This difference is because the Eurozone still has some structural problems and inflation is affected by many different global and local causes besides just sudden changes in money supply. The study shows that monetary policy’s transmission routes were still working even when standard interest rate tools weren’t working. In the end, this study adds to the growing amount of proof that unconventional tools are strategically important in modern central banking. However, this study shows that quantitative easing works in Europe, especially when it comes to boosting economic activity when interest rates are zero. However, it also says that these kinds of policies aren’t always effective at keeping inflation in check, which suggests that fiscal and structural changes may be needed together to achieve overall macroeconomic stability. 3. Conclusion This literature review examines the European Central Bank’s (ECB) monetary policy from its inception, focusing on its theoretical foundations, strategic development, and empirical evaluations. The primary responsibility of the ECB is to maintain price stability. This concept is integrated throughout the firm and implemented through a strategy that amalgamates financial and economic analysis. Works by Issing (2000) and Çolakoğlu (2004) provide a comprehensive analysis of the initial segment of the ECB’s policy. This section emphasized a structured, rules-oriented system with explicit inflation objectives and benchmark values for monetary aggregates. These strategies originated from monetarist concepts, although they were adapted to align with the Eurozone’s intricate institutional framework. Taylor (1999) and Sauer & Sturm (2007), show that straightforward interest rate rules, particularly Taylor-type rules, might assist individuals in making steady and lucid financial decisions, even amongst economic fluctuations and uncertainties. However, Bibow (2002) criticized the ECB for its cautious approach and conservative 46   RESEARCH, METHODS AND ANALYSIS IN SOCIAL SCIENCES . . . responses to inflation fluctuations, even at the expense of deteriorating actual economic performance. Cengiz (2019) asserts that the ECB’s shift towards more transparency has significantly influenced the efficacy of policies, particularly in contexts where conventional instruments such as interest rate adjustments were constrained. In addition to the significance of the ECB’s official policy instruments, its capacity to manage market expectations and provide forward guidance through organized communications, speeches, and press conferences was as crucial. Nonetheless, several studies indicated that this communication was not clear nor consistent, adversely affecting the bank’s image and prompting an erroneous market reaction. The global financial crisis of 2007–2008 and the European debt crisis that followed were big turning points in ECB policy. Due to the constraints of conventional monetary instruments, the ECB employed an array of unconventional measures, including long-term refinancing operations and extensive asset purchase programs. Claeys & Demertzis (2017), Claeys et al. (2018), and Zabala & Prats (2020) demonstrate that these unconventional measures contributed to economic stability, mitigated deflation, and sustained elevated aggregate demand as interest rates approached zero. However, they also introduced additional challenges, such as managing the central bank’s balance sheet size, devising effective exit strategies, and addressing the political complexities associated with purchasing national bonds and coordinating the budget. The analysis indicates that the ECB faces a fundamental tension: it must uphold its commitment to price stability while simultaneously responding swiftly and efficiently to evolving economic shocks. This is particularly significant in a currency union without fiscal authority or political cohesion, as economic disparities across member states can exacerbate the uneven impacts of uniform monetary policies. Surico (2007) and Garcia-Iglesias (2007) indicate that the European Central Bank’s actual policy implementation frequently exhibits nonlinearities and asymmetric responses, favoring stringent inflation control above growth assistance. This indicates that the ECB is more prudent than previously assumed. The ECB’s monetary policy necessitates a meticulous equilibrium between tradition and innovation due to the evolving economic landscape. While preserving price stability remains the primary objective of its mandate, increased flexibility and adaptation are necessary due to the growing complexity of financial markets, persistent low inflation, and emerging geopolitical risks. The ECB’s ability to enhance cooperation with fiscal authorities across member THE MONETARY POLICY OF THE EUROPEAN CENTRAL BANK . . .   47 states, implement data-driven decision-making, and refine its communication strategies will likely dictate the efficacy of its policies moving forward. Comprehending the historical evolution, strategic adjustments, and critical evaluations of the ECB’s monetary policy elucidates its previous achievements and shortcomings, as well as the institution’s potential trajectory in addressing forthcoming economic challenges. References Akbakay, Z. (2018). The Monetary Policy of the European Central Bank: Quantitative Definition of Price Stability and The Two Pillar Analysis Approach. Balkan Sosyal Bilimler Dergisi, 7(14). Bibow, J. (2002). The Monetary Policies of the European Central Bank and the Euro’s (Mal-) Performance: a stability-oriented assessment. International Review of Applied Economics, 16(1), 31-50. Cengiz, V. (2019). Merkez bankalarının iletişimi ve para politikası: Avrupa Merkez Bankası örneği. Balkan Sosyal Bilimler Dergisi, 8(15), 53-59. Claeys, G., & Demertzis, M. (2017). How should the European Central Bank ‘normalise’its monetary policy? (No. 2017/31). Bruegel Policy Contribution. Claeys, G., Demertzis, M., & Mazza, J. (2018). A monetary policy framework for the European Central Bank to deal with uncertainty (No. 2018/21). Bruegel Policy Contribution. Çolakoğlu, B. (2004). Avrupa Merkez Bankası’nın Para Politikası Stratejisi ve Teorik Temelleri. Bilgi Sosyal Bilimler Dergisi, (8), 95-112. Garcia-Iglesias, J. M. (2007). How the European Central Bank decided its early monetary policy?. Applied Economics, 39(7), 927-936. Issing, O. (2000). The monetary policy of the European Central Bank: strategy and implementation; the European Monetary Union. In CESifo Forum (Vol. 1, No. 2, pp. 3-9). München: ifo Institut für Wirtschaftsforschung an der Universität München. Sauer, S., & Sturm, J. E. (2007). Using Taylor rules to understand European Central Bank monetary policy. German Economic Review, 8(3), 375-398. Surico, P. (2007). The monetary policy of the European Central Bank. Scandinavian Journal of Economics, 109(1), 115-135. Taylor, J. B. (1999). The robustness and efficiency of monetary policy rules as guidelines for interest rate setting by the European Central Bank. Journal of Monetary Economics, 43(3), 655-679. 48   RESEARCH, METHODS AND ANALYSIS IN SOCIAL SCIENCES . . . Zabala, J. A., & Prats, M. A. (2020). The unconventional monetary policy of the European Central Bank: Effectiveness and transmission analysis. The World Economy, 43(3), 794-809. UNEMPLOYMENT AND EDUCATION EXPENDITURES: TESTING FOR . . .   55 education expenditures have a cyclical relationship, with each influencing the other. Öksüz (2023) examined the relationship between public education expenditure and youth unemployment rates in Turkey for women and men between 1988 and 2020, employing the Fourier–Granger causality test. The study's findings suggest a causal link between spending on education and female youth unemployment rates. No causal link was discovered between youth female unemployment and education spending; between education spending and youth male unemployment; or between youth male unemployment and education spending. For the period 2007-2021, Atay (2024) examined the causal link between public education expenditure and the elderly unemployment rate in 26 OECD countries. Analysis results showed a two-way causality relationship exists between the elderly unemployment rate and education expenditure. Studies in the literature suggest that rises in public expenditure on education can lead to a decrease in unemployment rates. This finding suggests that education spending has an influence on unemployment figures, especially when government spending on education aligns with labor market policies. 3. Research Method This study aims to explore the link between public expenditure on education and unemployment rates in developed EU member states within the OECD over the period from 2000 to 2021. Details concerning the dataset to be employed in the analysis and the methodology to be utilised will be provided below. 3.1. Data Set Data on public education expenditure as a share of GDP1 and unemployment rates for 202 developed European Union member countries within the OECD, covering 2000 to 2021, was obtained from the World Bank database. These 20 countries were selected for the period under consideration due to the continuity of the data. STATA18 statistical software was used for data analysis in this study. Table 1 provides information on the dataset used in the model. 1 General government expenditure on education (current, capital and transfers) is expressed as a percentage of total general government expenditure across all sectors (including health, education, social services, etc.). It includes expenditure financed by transfers to the government from international sources. 2 Germany, Austria, Belgium, Czech Republic, Denmark, Estonia, Finland, France, Netherlands, Spain, Sweden, Italy, Latvia, Lithuania, Hungary, Poland, Portugal, Slovakia, Slovenia, Greece. 56   RESEARCH, METHODS AND ANALYSIS IN SOCIAL SCIENCES . . . Table 1. Data Set Abbreviation Variables Sources unemployment Unemployment Rate World Bank education Share of Public Education Expenditure in GDP World Bank The equation relating to the model used in the study is as follows. ( ) #1 it it unemployment education u=+ The unemployment variable in the study›s model denotes the yearly unemployment rate across various countries, the education variable denotes the yearly proportion of education expenditure in a country’s GDP, and the u variable represents the error term. 3.2. Econometric Method The study initially investigated cross-sectional dependence. Crosssectional dependence between series in panel data analysis significantly impacts the results obtained (Pesaran, 2004). Pesaran’s (2004) cross-sectional dependence test can be applied in two different manners. The preferred cross-sectional dependence test is the one used when T and N are large. The test statistic was derived from the following equation: ( ) ( ) ( ) 1 2 2 11 1 # 1 2 1ˆ NN ij i ji CDLM T NN r - = =+ =- -åå Pesaran’s (2004) second test statistic is used when N>T. The calculation method for this test statistic is as follows. ( ) ( ) 1 11 2 ˆ #3 1 NN ij i ji T CDLM NN r - = =+ æö =ç÷ ç÷ -èø åå The Pesaran (2004) test can be applied to unbalanced panels as well as balanced panel models. The equation developed by Pesaran (2004) for unbalanced panels is as follows: ( ) ( ) 1 11 2 #4 ˆ 1 NN ij ij i ji CD T NN r - = =+ æö =ç÷ ç÷ -èø åå UNEMPLOYMENT AND EDUCATION EXPENDITURES: TESTING FOR . . .   57 The next step in panel data analysis is to conduct a homogeneity or heterogeneity test. This study utilised the Delta test developed by Pesaran and Yamagata (2008) as its homogeneity test. This test investigates whether the slope coefficients have homogeneous values (Pesaran and Yamagata, 2008, s. 71). The hypotheses of the Delta test are outlined as follows: H0: The slope coefficient is homogeneous. H1: The slope coefficient is heterogeneous. The multivariate augmented Dickey-Fuller (MADF) test was proposed by Sarno and Taylor in 1998. This test allows for simultaneous cross-sectional correlations among panel members (Lau, 2009, s. 5). The MADF test has two principal benefits. The test is initially based on a higher-order autoregressive equation, as opposed to a first-order autoregressive equation. In particular, the MADF test does not enforce the condition that autoregressive coefficients have to be uniform across all panel members; it allows for varying autoregressive coefficients across individual panel members (Furuoka, 2012, p. 134). The only conclusion one can draw if the MADF test rejects the null hypothesis is that at least one panel member has a stationary root (Breuer, 2001, s. 488). Taylor and Sarno (1998) formulate the hypotheses of the MADF test as follows. H0: At least one series in the panel is non-stationary. H1: All series in the panel are stationary. After conducting unit root tests, Westerlund (2007) used co-integration analysis to determine whether a long-term reciprocal relationship existed between the series. Westerlund’s (2007) panel co-integration tests, designed to examine the long-term co-integration relationship between the dependent and independent variables in a panel data set, are commonly used tests in empirical studies. The Westerlund (2007) co-integration test, which is based on the error correction model, comprises four test statistics. These studies include two that investigate co-integration in cross-sections for the Ga and Gt group mean statistics, as well as Pa and Pt that examine co-integration in the panel. The equation proposed by Westerlund (2007) is provided in the following (Westerlund, 2007, pp. 717-718). ∑  ( )  ( ) ∑  ( ) ( )  ( )  (  ) ( ) 58   RESEARCH, METHODS AND ANALYSIS IN SOCIAL SCIENCES . . . ∑  ( )  ( ) ∑  ( ) ( )  ( )  (  ) ( ) Westerlund (2007) defined the following hypotheses for group statistics and statistics for the entire panel in the co-integration test: H0: There is no co-integration between the series. H1: There is co-integration between the series. Rejecting the null hypothesis (H0) for Westerlund’s (2007) test suggests that there is a long-term relationship between the variables across the entire panel, whereas accepting the H0 hypothesis implies that there is no long-term relationship between the variables across the entire panel. The test used to establish a causality relationship between variables in panel data analysis is the Dumitrescu and Hurlin (2012) causality test. Given that a causality relationship valid for one country in terms of an economic phenomenon is also likely to be valid for other countries, the Dumitrescu Hurlin panel causality test allows for more effective testing of causality relationships within the panel data framework with more observations (Bozoklu and Yılancı, 2013, s. 175). Dumitrescu and Hurlin’s panel causality analysis offers several advantages: it recognises cross-sectional dependence among units in the panel, considers variability in the slope coefficients across units, can be applied when the time dimension is either smaller or larger than the cross-sectional dimension, and can produce effective results even with unbalanced panel data sets (Dumitrescu and Hurlin, 2012, p. 1457). Dumitrescu and Hurlin (2012) used the following linear equation to model the causality relationship between X and Y (Dumitrescu and Hurlin, 2012: 1451): ( ) ( ) ( ) 11 #9 KK kk it i it k it k it kk y a gg bg e -- == =+ + + åå The following hypotheses are tested in the Dumitrescu and Hurlin (2012) method: H0: The y variable is not causal for the x variable for all units. H1: The y variable is causal for the x variable for some units. UNEMPLOYMENT AND EDUCATION EXPENDITURES: TESTING FOR . . .   59 These hypotheses will be tested using the Wald statistic. The formula for the mean of this statistic, , HnC NT W , is as follows (Dumitrescu and Hurlin, 2012, s. 1453): ( ) ,, 1 1. # 10 N HnC NT it i WW N = =å 3.3. Findings The presence of cross-sectional dependence in the variables was initially assessed using Pesaran’s (2004) CD test. The results are shown in Table 2. Table 2. Cross-sectional Dependency Test Variables CD-test Probability Value unemployment 18.00 0.000* education 8.41 0.000* Note: * indicate statistical significance at the %1 significance levels respectively. Evaluating the probability values in Table 2 at a 1% significance level shows that the null hypothesis is rejected, leading to the conclusion that there is cross-sectional dependence in the model. A shock that happens in the countries being analysed will also impact other countries. If cross-sectional dependence exists in the panel data analysis, the stationarity of the series is then tested using a second-generation panel unit root test. The homogeneity of the model’s slope coefficient across cross-sectional units was then examined using the Delta Test developed by Pesaran and Yamagata (2008) after conducting a test for cross-sectional dependence in the model. The results of the Delta Test are detailed in Table 3. Table 3. Homogeneity Test Variables Test Statistic Probability Value Delta 9.963 0.000* Delta adj. 10.721 0.000* Note: * indicate statistical significance at the %1 significance levels respectively. The data in Table 3 indicate a statistically significant calculated probability value at 1% significance level, leading to the rejection of the null hypothesis and the categorisation of the model as heterogeneous. 60   RESEARCH, METHODS AND ANALYSIS IN SOCIAL SCIENCES . . . Due to the cross-sectional dependence between series in the unit root test, second-generation panel unit root tests are required. The multivariate augmented Dickey-Fuller (MADF) unit root test, accounting for cross-sectional dependence, was employed in this study. The results of the MADF test, one of the second-generation unit root tests, are illustrated in Table 4. Table 4. Unit Root Test Variables Lags MADF Approx unemployment 1 1258.196 36.616 education 1 6340.197 36.616 The calculated MADF test statistic is 1258.196 for the unemployment variable and 6340.297 for the education variable. Since these values are greater than the critical value of 36.616, the null hypothesis is rejected and the variables are determined to be stationary at the level. A co-integration test conducted by Westerlund in 2007 aimed to determine if a co-integration relationship existed between the model’s variables. Results of the Westerlund (2007) co-integration test are shown in Table 5. Table 5. Cointegration Test Statistic Value Z-value P-value Gt -2.184 -1.967 0.025** Ga -9.572 -1.957 0.025** Pt -11.187 -4.533 0.000* Pa -8.309 -3.853 0.000* Note: *, ** indicate statistical significance at the 1%, 5% significance levels respectively. The Westerlund Cointegration Test results in Table 5 indicate rejection of the null hypothesis (H0) since the probability values for the Ga and Gt statistics, and the Pa and Pt statistics are all less than 0.05. The results show that there is a long-term relationship between the variables. The existence of a causal relationship between public education expenditure and unemployment was examined using the panel causality method developed by Dumitrescu and Hurlin (2012). The final stage of the analysis process yielded the results of the Dumitrescu and Hurlin (2012) panel causality test, which are UNEMPLOYMENT AND EDUCATION EXPENDITURES: TESTING FOR . . .   61 presented in Table 6. The Wald, asymptotic Z, and semi-asymptotic Z statistics from the estimates are summarised in the table. Table 6. Dumitrescu and Hurlin (2012) Panel Causality Test Estimation Results Models W-bar Statistic Z-bar Probability Value Z-bar tilde Probability Value education to unemployment 2.5318 4.8441 0.0000* 3.5886 0.0003* unemployment to education 2.2016 3.7998 0.0001* 2.7463 0.0060* Note: * indicate statistical significance at the %1 significance levels respectively. According to the Dumitrescu and Hurlin causality test results shown in Table 6, a two-way causal relationship was identified, running both from public education expenditures to unemployment and from unemployment to public education expenditures. Determining this relationship allows us to conclude that the unemployment rate and the education expenditures-to-GDP ratio, as well as the education expenditures-to-GDP ratio, are related to the unemployment rate in the countries under study during the relevant period. Based on this result, we can conclude that education expenditures play an effective role in reducing unemployment. 4. Conclusion Many nations are grappling with economic and social problems resulting from unemployment. Governments in different countries use a range of policies to tackle the issue of unemployment. Enhancing the quality of educational services is one of these policies. Developing knowledge, skills and abilities through education enables individuals to enter the labor market, ultimately lowering their risk of unemployment. Within this context, education spending holds a significant position among public sector expenditures. Education’s significant role in various sectors underscores the importance of its impact on economic growth and stability. This study seeks to investigate the causal link between education expenditures and unemployment rates in 20 developed EU member states, which are OECD members, from 2000 to 2021, using the Dumitrescu-Hurlin causality test. The 62   RESEARCH, METHODS AND ANALYSIS IN SOCIAL SCIENCES . . . analyses uncovered a two-way causal relationship between unemployment and education expenditures. This finding is in line with the research conducted by Han (2021), Arıca and İpsal (2021)3, Yardımcı (2022), Doğan (2023), Öksüz (2023)4, and Atay (2024)5 in previous studies. Studies have shown that there is a two-way relationship between education spending and unemployment levels, with education spending affecting unemployment rates and unemployment levels influencing education spending. Increased expenditure in higher education can result in improved educational services, broader access to education for a greater number of individuals, and an overall improvement in educational standards within a community. Therefore, with higher educational levels among individuals, their job prospects and earning potential may both be enhanced, ultimately leading to lower unemployment rates. Elevating educational standards and reducing unemployment rates are of the most importance in policy terms. But unemployment rates are also influenced by a country’s overall macroeconomic policies as well as its education policies. In countries where unemployment is rising, macroeconomic indicators will largely fail to give positive signals. Public policies will be shaped by a variety of factors, with a focus on achieving macroeconomic goals within the context of creating and enhancing employment opportunities. Rising unemployment has significant implications for public policy, potentially resulting in increased spending on higher education. Countries put in place diverse public education policies to improve educational standards, promote fair opportunities in education, and eliminate barriers to equal access. It affects education spending within the budget and may result in increased spending. References Arıca, F. ve İpsal, S. (2021). Eğitim harcamaları-genç işsizlik nedensellik ilişkisi: Seçilmiş OECD ülkeleri için panel veri analizi. Girişimcilik ve Kalkınma Dergisi, 16(2), 16-30. Atay, E. (2024). Yaşlı işsizliği ve kamu eğitim harcamaları ilişkisinin Dumitrescu-Hurlin panel nedensellik testi ile belirlenmesi. Toplumsal Politika Dergisi, 5(2), 121-137. 3 The relationship between education expenditure and youth unemployment has been examined. 4 The relationship between education expenditure and youth unemployment among women and men has been examined. 5 The relationship between education expenditure and unemployment among older people has been examined. UNEMPLOYMENT AND EDUCATION EXPENDITURES: TESTING FOR . . .   63 Binuomoyo, O. K. (2020). Examining the relationship between public spending on education and unemployment problem in Nigeria. Malaysian Journal of Business and Economics, 7(1), 57. Bozuklu, Ş. ve Yılancı, V. (2013). Finansal gelişme ve iktisadi büyüme arasındaki nedensellik ilişkisi: Gelişmekte olan ekonomiler için analiz. Dokuz Eylül Üniversitesi İktisadi ve İdari Bilimler Fakültesi Dergisi, 28(2), 161-187. Breuer, J. B., McNown, R., & Wallace, M. S. (2001). Misleading inferences from panel unit‐root tests with an illustration from purchasing power parity. Review of International Economics, 9(3), 482-493. Çalışkan, Ş. (2007). Eğitim-işsizlik ve yoksulluk ilişkisi. Sosyal Ekonomik Araştırmalar Dergisi, 7(13), 284-308. Çevik, A., & Yüksel, C. (2021). Yarı kamusal mallar ve ssimetrik bilgi arasındaki ilişki bağlamında sağlık hizmetleri. Dicle Üniversitesi İktisadi ve İdari Bilimler Fakültesi Dergisi, 11(21), 85-107. Çondur, F., & Cömertler Şimşir, N. (2017). Türkiye’de eğitim harcamaları, ekonomik büyüme ve genç işsizlik ilişkilerinin analizi. Uluslararası Bilimsel Araştırmalar Dergisi, 2(6), 44-59. Dachito, A. C., Alemu, M., & Alemu, B. (2020). The impact of public education expenditures on graduate unemployment: Cointegration analysis to Ethiopia. Journal of International Trade, Logistics and Law, 6(2), 62-78. Doğan, M. (2023). Kamu eğitim harcamalarının gelir eşitsizliği ve istihdamla ilişkisinin panel veri yöntemiyle analizi. Maliye Dergisi, 184, 171196. Dumitrescu, E. I., & Hurlin, C. (2012). Testing for Granger non-causality in heterogeneous panels. Economic Modelling, 29(4), 1450-1460. Filho, N. A. M. (2013). Perspectives on labour economics for development (Cazes, S. & Verick, S. Ed.) Geneva, Switzerland: International Labour Organization. Furuoka, F. (2012). Unemployment hysteresis in the East Asia‐Pacific region: new evidence from MADF and SURADF tests. Asian‐Pacific Economic Literature, 26(2), 133-143. Göker, Z. (2008). Kamusal mallar tanımında farklı görüşler. Maliye Dergisi, 155(2), 108-118. Hall, J. (2000). Investment in education: Private and public returns. West Virginia University Department of Economics Working Paper Series, Working Paper No. 16-05, 64   RESEARCH, METHODS AND ANALYSIS IN SOCIAL SCIENCES . . . Han, V. (2021). The impact of technological growth and education spending on unemployment: Evidence from a panel ARDL-PMG approach. Journal of Entrepreneurship and Innovation Management, 10(2), 1-22. Horner, S., Zhang, A., & Furlong, M. (2018). The impact of higher education on unemployment. https://smartech. gatech. edu/bitstream/handle/1853/60545/ econ_3161_research_paper. pdf. Işık, S. (2020). Küresel rekabet açısından yatırım carilerinin önemi: Türkiye üzerine bir analiz. Yönetim Bilimleri Dergisi, 18(35), 9-35. https://doi. org/10.35408/comuybd.468255 Lau, C. K. M. (2009). A more powerful panel unit test with an application to PPP. Available at SSRN 1347942, 1-14. Maral, M., Yıldız, F. ve Alpaydın, Y. (2021). Türkiye’de yüksek öğretim harcamaları ve genç işsizliği ilişkisi üzerine bir analiz. Journal of Economic Policy Researches, 8(2), 173-197. Mehmetaj, N. ve Xhindi, N. (2022). Public expenses in education and youth unemployment rates—A vector error correction model approach. Economies, 10(12), 293, Onuoha, F. C., & Agbede, M. (2019). Impact of disagregated public expenditure on unemployment rate of selected African countries: A panel dynamic analysis. Journal of Economics, Management and Trade, 24(5), 1-14. Öksüz, M. (2023). Eğitim harcamaları ile cinsiyete göre genç işsizlik ilişkisi: Türkiye örneği. Yönetim Bilimleri Dergisi, 21(49), 459-481. https://doi. org/10.35408/comuybd.1250248 Pata, U. K. (2020). OECD ülkelerinde işsizlik histerisinin ampirik bir analizi: Fourier panel durağanlık testi. Sosyal Güvenlik Dergisi, 10(1), 125-144. Pesaran, M. H. (2004). General diagnostic tests for cross section dependence in panels. Cambridge Working Papers in Economics, 1240(1), 1-39. Pesaran, M.H. ve Yamagata, T. (2008). Testing slope homogeneity in large panels. Journal of Econometrics, 142(1), 50–93. Erişim adresi: https:// www. sciencedirect.com/science/article/pii, Erişim Tarihi: 14.08.2024. Pirim, Z., Owings, W. A., & Kaplan, L. S. (2014). The long-term impact of educational and health spending on unemployment rates. European Journal of Economic and Political Studies, 7(1), 49-69. Selase, A. E. (2019). Impact of disaggregated public expenditure on unemployment rate of selected african countries: A panel dynamic analysis approach. American International Journal of Humanities, Arts and Social Sciences, 1(2), 47-57. HOW TO DEAL WITH MISSING OBSERVATIONS IN PANEL DATA . . .   71 As you can see from Figure 3 above, msp variable has 16 missing observations out of 28534 observations, and collgrad has zero missing variables. All other variables can be checked with this command. Also, be aware that when using reg y x1 x2 (.e. basic regression ) and xtreg y x1 x2, fe commands, both automatically excludes missing observations in the data without doing anything. In order to check which one is exactly missing, we can use the command: .list if e(sample)==0 , after running the regression. To generate a list of missing values by panel, please utilize the following command: .bysort panelid (timevar): gen miss_x = missing(x) In instances where there is a considerable amount of missing data, it is advisable to consider omitting specific values, particularly if a variable is entirely absent or nearly completely missing for certain panel identifiers or time periods. A clearer understanding of this situation can be attained by utilizing the commands provided in the list below: Obviously, if there are too much missing, one should consider to omit values from the data especially if variable is missing entirely or close to the entire for certain panelid or time period. There is a way of understanding this by using the list of command below: .egen tag = tag(panelid) .bysort panelid (timevar): gen allmiss = missing(x) .bysort panelid: egen totalmiss = total(allmiss) .drop if totalmiss == _N // drop entire panel with full missing This information is particularly beneficial when managing large datasets, as it is impractical to examine each entry individually. Additionally, another command that can be utilized is `xtdescribe`. It is important to exercise caution with commands that begin with “xt,” as they may only be executed if the dataset has been configured as panel data. Please refer to Figure 4 below for the output generated by this command: 72   RESEARCH, METHODS AND ANALYSIS IN SOCIAL SCIENCES . . . Figure 4. The Output of xtdescribe command As illustrated in Figure 4 above, the panel showcases imbalanced patterns across the dataset. This observation provides valuable insights into the underlying data patterns. An additional useful command is ‘xtsum.’ Figure 5 below presents the overall descriptive statistics, and, in conjunction with the ‘summarize’ command, it offers separate standard deviations for both between-group and within-group analyses. This functionality facilitates the identification of any missing observations when comparing the variables against one another. HOW TO DEAL WITH MISSING OBSERVATIONS IN PANEL DATA . . .   73 within .29266 -.4077221 4.78367 T-bar = 6.05689 between .424569 0 3.912023 n = 4711 ln_wage overall 1.674907 .4780935 0 5.263916 N = 28534 within 23.96999 -18.43924 131.156 T-bar = 5.93918 between 20.64508 0 104 n = 4686 wks_work overall 53.98933 29.03232 0 104 N = 27831 within 7.520712 -2.154726 130.0596 T-bar = 6.04395 between 7.846585 1 83.5 n = 4710 hours overall 36.55956 9.869623 1 168 N = 28467 within 2.659784 -14.27894 15.62384 T-bar = 5.98021 between 2.796519 0 21.16667 n = 4699 tenure overall 3.123836 3.751409 0 25.91667 N = 28101 within 3.484133 -9.642671 20.38091 T-bar = 6.05689 between 3.724221 0 24.7062 n = 4711 ttl_exp overall 6.215316 4.652117 0 28.88461 N = 28534 within 6.054 -33.95191 64.38143 T-bar = 4.91496 between 5.181437 0 76 n = 4645 wks_ue overall 2.548095 7.294463 0 76 N = 22830 within .2668622 -.6822348 1.151099 T-bar = 4.63566 between .3341803 0 1 n = 4150 union overall .2344319 .4236542 0 1 N = 19238 within 1.650248 -5.522328 15.44434 T-bar = 6.04661 between 2.86512 1 13 n = 4699 occ_code overall 4.777672 3.065435 1 13 N = 28413 within 1.708429 -1.507027 17.12154 T-bar = 6.0049 between 2.542844 1 12 n = 4695 ind_code overall 7.692973 2.994025 1 12 N = 28193 within .1597932 -.5237771 1.34289 T-bar = 6.05519 between .4667982 0 1 n = 4711 south overall .4095562 .4917605 0 1 N = 28526 within .2490022 -.5761154 1.290551 T-bar = 6.05519 between .4271586 0 1 n = 4711 c_city overall .357218 .4791882 0 1 N = 28526 within .1834446 -.6461273 1.215777 T-bar = 6.05519 between .4111053 0 1 n = 4711 not_smsa overall .2824441 .4501961 0 1 N = 28526 within 0 .1680451 .1680451 T-bar = 6.05689 between .4045558 0 1 n = 4711 collgrad overall .1680451 .3739129 0 1 N = 28534 within 0 12.53259 12.53259 T-bar = 6.05904 between 2.566536 0 18 n = 4709 grade overall 12.53259 2.323905 0 18 N = 28532 within .2456558 -.7036538 1.163013 T-bar = 6.05349 between .3684416 0 1 n = 4711 nev_mar overall .2296795 .4206341 0 1 N = 28518 within .3238927 -.3304159 1.536251 T-bar = 6.05349 between .3982385 0 1 n = 4711 msp overall .6029175 .4893019 0 1 N = 28518 within 0 1.303392 1.303392 T-bar = 6.05689 between .4862111 1 3 n = 4711 race overall 1.303392 .4822773 1 3 N = 28534 within 5.16945 14.79511 43.79511 T-bar = 6.05308 between 5.485756 14 45 n = 4710 age overall 29.04511 6.700584 14 46 N = 28510 within 0 48.08509 48.08509 T-bar = 6.05689 between 3.051795 41 54 n = 4711 birth_yr overall 48.08509 3.012837 41 54 N = 28534 within 5.138271 63.79198 92.70865 T-bar = 6.05689 between 5.156521 68 88 n = 4711 year overall 77.95865 6.383879 68 88 N = 28534 within 0 2601.284 2601.284 T-bar = 6.05689 between 1487.57 1 5159 n = 4711 idcode overall 2601.284 1487.359 1 5159 N = 28534 Variable Mean Std. Dev. Min Max Observations . xtsum Figure 5. The output of xtsum command 74   RESEARCH, METHODS AND ANALYSIS IN SOCIAL SCIENCES . . . An additional command that proves to be useful is outlined below. This command enables the user to accurately count the number of missing variables in a specified order. .foreach var of varlist var1 var2 var3 { count if missing(`var’) } The analysis presented in Figures 6 and 7 indicates that the variables age, race, msp, and grade exhibit 24, 0, 16, and 2 missing values, respectively. In contrast, the variables tenure and hours demonstrate 433 and 67 missing observations. 2 16 0 24 3. } 2. count if missing(`var') . foreach var of varlist age race msp grade { Figure.6. Output with four variables 67 433 16 0 24 . } 3. . count if missing(`var') 2. . foreach var of varlist age race msp tenure hours { Figure.7 Output with five variables The command .list var1 var2 if missing(var1) | missing(var2) may be utilized to identify any missing observations present in one or both of the specified variables within the dataset. A sample output illustrating the results of this command is provided in Figure 8. Finally, we can use the command list var1 var2 if missing (var1) | missing (var2) for either one or more variables in the data set to see if any HOW TO DEAL WITH MISSING OBSERVATIONS IN PANEL DATA . . .   75 missing observation is present. Figure 8.a and 8.b are sample outputs for this command. . 28449. . 16 28117. 41 . 25765. . 12 25588. . 12 25236. . 9 24132. . 12 22089. . 12 19660. . 9 18199. . 11 17999. . 15 16140. . 10 15546. . 11 15333. . 14 14322. . 16 14247. . 8 13524. . 12 13506. . 14 13491. . 12 11739. . 10 10946. . 12 10375. . 12 9235. . 12 6356. 37 . 3574. . 13 2026. . 15 1676. . 14 age grade . list age grade if missing(age) | missing(grade) . 28449. . 16 black 28117. 41 . white 25765. . 12 black 25588. . 12 black 25236. . 9 white 24132. . 12 black 22089. . 12 white 19660. . 9 white 18199. . 11 white 17999. . 15 white 16140. . 10 white 15546. . 11 black 15333. . 14 black 14322. . 16 black 14247. . 8 black 13524. . 12 black 13506. . 14 white 13491. . 12 white 11739. . 10 white 10946. . 12 white 10375. . 12 white 9235. . 12 white 6356. 37 . black 3574. . 13 white 2026. . 15 white 1676. . 14 white age grade race . list age grade race if missing(age) | missing(grade) | missing(race) Figure 8. A Sample Output for list var1 var2 if missing(var1) | missing(var2) command in STATA 8.a. With two variables 8.b. With three variables This command not only provides an overview of the number of missing variables but also specifies the precise locations of the missing observations. This information is essential in determining the appropriate course of action regarding the missing data, whether to omit these entries or to apply a suitable method for imputation. Identifying missing observations is fundamentally important for making informed decisions on how to address these issues in subsequent analyses. In the context of panel data analysis, it is preferable to utilize a balanced panel rather than an imbalanced one, as the latter may result in inefficiencies and potential biases. Consequently, it is vital to address missing observation issues, particularly when the absence of data is not random or when it correlates with the dependent variable. Furthermore, there are cases in which models necessitate complete data, such as individual-level survey data (Balogun 76   RESEARCH, METHODS AND ANALYSIS IN SOCIAL SCIENCES . . . O.P. et al. 2022; Brygala, M. 2022; Castaneda Rodriguez 2018; Jauch, S. & Watzka 2016). 4. Dealing with missing observation: Dropping, Replacing or Imputation 4.1 Dropping: A widely recognized and effective strategy for managing missing values within a dataset is to eliminate these entries. This approach contributes to the overall integrity of the data analysis while ensuring that the findings remain both valid and reliable. When you type; . drop if missing (varname) The aforementioned figure illustrates that the variable “tenure” contains 433 missing observations. In order to demonstrate the functionality of the subsequent command, we will examine the output, which will automatically eliminate the missing values associated with this specific variable. (433 observations deleted) . drop if missing(tenure) As indicated in the preceding section, a total of 433 observations have been removed. Nevertheless, there exists a more efficient method to eliminate all missing observations with a single command, as illustrated by the following command: .drop if missing(var1, var2, var3, etc.) (9,647 observations deleted) > l_exp, tenure, hours, wks_work) . drop if missing( birth_yr, age, race, msp, nev_mar, grade, collgrad, not_smsa, c_city, south, ind_code, occ_code, union,tt This methodology will automatically reduce the sample size when addressing missing values, which is particularly prevalent when these values are situated either at the beginning or the end of the dataset. For instance, it is possible that 9,647 observations may be removed. Once the missing observations have been eliminated using this command, users should repeat the command to verify whether any missing values remain. Upon doing so, it should be evident that there are no outstanding missing values. HOW TO DEAL WITH MISSING OBSERVATIONS IN PANEL DATA . . .   77 (0 observations deleted) > tenure, hours, wks_work) . drop if missing( birth_yr, age, race, msp, nev_mar, grade, collgrad, not_smsa, c_city, south, ind_code, occ_code, union,ttl_exp, (9,647 observations deleted) > l_exp, tenure, hours, wks_work) . drop if missing( birth_yr, age, race, msp, nev_mar, grade, collgrad, not_smsa, c_city, south, ind_code, occ_code, union,tt No more missing observation is left in the panel data. (i.e. zero observation is deleted). Another common case is when the missing observations are in between. This is typically very common issue with the panel data to have missing observations scattered around the time period which in most cases create a problem of ending up with un unbalanced panel instead of balanced panel. This leads to have an additional diagnostic test and in most cases it either makes the diagnostics tests more demanding or we end up with lack of efficiency and or biasness. Before exploring this any further, perhaps one should bear that in mind that if assuming there is a continues function between closed interval, in this case it is a time period (i.e. [t1,t2] t for time) , and differentiable over the open interval (i.e. (t1,t2), then there exist a point ti in this interval (t1,t2), such that the tangent line to the graph of function at ti is parallel to the secant line connecting (t1,f(t1)) and (t2, f(t2)). (Mean value theorem, (Tirado-Serrato, 2023)). Which tells us in a way that the missing values when scattered around the data, should have values that they lie within the time interval. Therefore, below is few methodology to handle such missing variables: A prevalent challenge in data analysis occurs when observations are missing at various points within the dataset. This issue is particularly common in panel data, where missing observations are often distributed throughout the time period, resulting in an unbalanced panel rather than a balanced one. Consequently, this imbalance necessitates additional diagnostic tests, which can complicate the analysis and potentially lead to inefficiencies or biases in the results. Before further exploring this issue, it is essential to acknowledge the assumption that if a continuous function is defined over a closed interval— specifically, the time period denoted as [t1, t2]—and is differentiable over the open interval (t1, t2), there exists a point ti within (t1, t2) such that the tangent line to the graph of the function at ti is parallel to the secant line connecting the points (t1, f(t1)) and (t2, f(t2)). This principle is articulated in the Mean Value Theorem (Tirado-Serrato, 2023). This theorem indicates that the missing values, when scattered throughout the dataset, should lie within the defined time interval. The following are various methodologies that have been proposed to tackle the issue of missing values: 78   RESEARCH, METHODS AND ANALYSIS IN SOCIAL SCIENCES . . . 4.2 Replacing: To maximize the retention of observations and thereby enhance the comprehensiveness of the dataset—avoiding any loss of years, variables, or individual observations—it is advisable to impute missing values by replacing them with the mean of the respective variable. First, the mean value of the variable is calculated with a command: .egen mean_x = mean(x) Upon calculating the mean value of the specified variable, we will replace all missing observations of this variable utilizing the command provided below: .replace x = mean_x if missing(x) For instance, the variable tenure exhibits a total of 433 missing observations (refer to Figure X above). We will proceed to execute the command pertaining to this variable and analyze the resulting outcome presented below: (433 real changes made) . replace tenure=mean_tenure if missing(tenure) . egen mean_tenure = mean(tenure) The 433 missing observations have been replaced with the mean value of the variable “tenure.” This approach can be replicated for all variables in which missing observations have been identified. While this method is not without its limitations, it aligns with the principles of the mean value theorem (TiradoSerrato, 2023). Additionally, another, albeit less conventional method involves substituting missing observations with zero, which can be executed using the appropriate command. .replace x = 0 if missing(x) Utilizing zero as a placeholder can facilitate the creation of a more comprehensive data set; however, this approach carries certain limitations. It is essential to recognize that zero represents only a single value and can exert a markedly different influence compared to the mean value. For instance, consider HOW TO DEAL WITH MISSING OBSERVATIONS IN PANEL DATA . . .   79 a scenario where there is a missing observation within a data set organized in a particular sequence: The dataset under consideration contains the following values: 3, _, 5, _, 8, _, _, _, _, 10, _, _, 14, _, _, _, _, where the underscores denote missing observations. In total, this dataset comprises 23 observations, of which 18 are identified as missing. The mean value of the available observations is 8. Applying the first replacement method, the missing values can be substituted as follows: 3 8 8 5 8 8 8 8 8 8 10 8 8 8 8 14 8 8 8 8 8 whereas, replacing them with 0 will turn the data into: 3 0 0 5 0 0 8 0 0 0 10 0 0 0 0 14 0 0 0 0 0 Both methodologies now appear to be complete; however, the descriptive statistics—and consequently their influence on the regression model—will exhibit considerable variation. Upon reviewing the mean value theorem (Tirado-Serrato, 2023), it is evident that the former methods are more justifiable. Furthermore, replacing missing values with zero may lead to additional complications when logarithmic transformations are applied, potentially resulting in the emergence of new missing values, if applicable. 4.3 Imputation Depending upon the structure of the panel data, or the missigness pattern, or the type of the variable, few imputation methods are used.The choice of imputation methods is contingent upon the structure of the panel data, the pattern of missing values, and the type of variable involved. In STATA the “mi” command is utilized for multiple imputation, which refers to the process of addressing missing observations within a dataset. This method offers an advanced approach to handling such issues. There are several vital commands associated with the “mi” functionality: When a dataset contains a significant proportion of missing values, it is imperative to first declare the dataset as a multiple imputation dataset using the relevant “mi” commands. Four common methods for achieving this are detailed below: The STATA has mi command that deals with multiple-imputation which is the abbreviation mi comes from the initials. It is an advanced way of handling the missing observation in the dataset, and there are few important command for mi below: 80   RESEARCH, METHODS AND ANALYSIS IN SOCIAL SCIENCES . . . To begin with, when there are missing values at an alarming level, first, the data set should be declared as mi. There are four common way of doing this through the commands below: .mi set wide .mi set mlong .mi set flong .mi set flongsep name It does not matter which one to start with as they can easily be converted to one another using the command: .mi convert By utilizing either command, we direct STATA to interpret the data as being multiple-imputed. An illustrative example can be found in Figure 9. Figure. 9 An output panel for mi set mlong command. When the dataset is configured for multiple imputation, the command `mi set mlong` generates several important variables: `_mi_miss`, which identifies the variables containing missing observations prior to imputation; mi_m, which ARTIFICIAL INTELLIGENCE-POWERED BRANDING IN THE METAVERSE . . .   87 with clear moral standards are more likely to earn people’s confidence over time and get ahead of their competitors in the digital marketplace (Sadiq & Devi, 2024). This chapter examines the fundamental concepts of AI-driven branding in the metaverse, emphasizing the impact of virtual reality on consumer shopping behavior. The chapter aims to give an overview of how firms could change to keep up with this fast-changing industry by looking at current trends, theoretical frameworks, and real-world examples. 2. Immersive Branding in the Metaverse The metaverse integrates the physical and digital realms to enhance customer experiences; nevertheless, there is limited understanding of consumer reactions to branded virtual environments (Wongkitrungrueng & Suprawan, 2024). Marketers, consumers, and scholars are demonstrating heightened interest in the metaverse, defined as computer-generated virtual worlds facilitating interaction and engagement among individuals. This immersive and interconnected digital environment integrates augmented reality, virtual reality, and the internet, allowing users to interact, cooperate, and engage in various activities within the metaverse. In this ecosystem, small-scale digital economies and virtual settlements have arisen, distinguished by non-fungible tokens (NFTs) that represent assets like virtual real estate, apparel, and vehicles. The metaverse’s economic architecture is founded on cryptocurrencies, which are secured by blockchain technologies (Simonetti et al., 2025). To thrive in today’s intensely competitive global and domestic markets, firms must go beyond mere customer satisfaction with superior products and services. The media and early metaverse companies have led the typical user to believe that in the metaverse, individuals utilize avatars for self-representation, engage with one another, and collaboratively construct a virtual community. Furthermore, digital currency is utilized in the metaverse to purchase clothing, artifacts, items in video games, and a diverse array of other products from various enterprises. Although it may initially appear to be science fiction, a more thorough analysis reveals that humans have been utilizing such technologies for an extended period (Chandiwala et al., 2023). Immersive technologies, such as augmented and virtual reality, gaming platforms, blockchain, and non-fungible tokens (NFTs), enable the integration of virtual and physical realities, signaling the anticipated advent of the metaverse. The metaverse is expected to signify the next advancement of the internet, 88   RESEARCH, METHODS AND ANALYSIS IN SOCIAL SCIENCES . . . envisioned as a more immersive alternative reality available to marketers and consumers. While a fully realized metaverse has not yet gained popular approval, the incorporation of immersive technologies is starting to emerge in the marketing strategies of fashion enterprises. Fashion brands must leverage emerging technologies to respond to changing consumer behavior and sustain a competitive edge (Chrimes & Boardman, 2023). Brand equity is one of the most valuable assets for modern organizations, and immersive brand experiences are becoming an important tool for strengthening it. As companies enter virtual environments, they can design distinctive interactions that leave a lasting impression on consumers. This shift requires businesses to adjust their strategies, as immersive technologies are likely to redefine how brand loyalty is built. To take advantage of this change, firms are investing in new technologies that support such experiences and allow them to connect with audiences on multiple sensory levels. In doing so, they not only reinforce their brand identity but also create meaningful relationships that encourage long-term customer commitment. Many companies are moving into the digital space out of fear of falling behind, but success will come only through approaches that are ethical, sustainable, and collaborative. In this setting, brands play an active role in creating long-term relationships with their audiences. By reshaping their strategies around changing customer expectations, firms can build loyalty and inspire advocacy (Catherine et al., 2024). 3. Fundamentals of AI-Powered Branding Brands are evolving through the adoption, implementation, and utilization of artificial intelligence (AI) technologies. AI fosters dynamic consumer-brand relationships that support the development of sustainable brands and long-term prosperity. Brands thrive in competitive markets by developing compelling brand narratives that comprehensively impact consumer purchasing decisions (Deryl et al., 2023). Emerging markets propel global economic growth through innovation to meet consumer demands. When consumers are satisfied, organizations prosper, and the entire market flourishes. These markets exhibit growing consumer demographics, increasing disposable incomes, and swift technology progress. The consumer experience is essential for transforming new customers into loyal patrons. Artificial intelligence is crucial for contemporary commercial performance, with those skilled in AI positioned for leadership roles. Artificial intelligence includes technologies such as machine learning, natural language ARTIFICIAL INTELLIGENCE-POWERED BRANDING IN THE METAVERSE . . .   89 processing, deep learning, and data analytics, which provide profound insights into customer behavior and preferences. By utilizing these technologies, brands may create tailored consumer experiences, improving customer satisfaction and meeting corporate objectives (Sekarini & Selvabaskar, 2024). Marketing and brand management employ numerous terms related to AI. These terms often highlight the innovative use of technology in understanding consumer behavior and optimizing campaigns. As brands increasingly leverage AI tools, they can create more personalized experiences that resonate with their target audiences. AI encompasses a multitude of functions and associated concepts, as it is frequently defined in a broad manner to incorporate various types of computers capable of directly executing or assisting tasks that formerly necessitated human emotions or cognition, utilizing software and algorithms. Consequently, research in marketing and branding frequently employs terminology such as “machine learning,” “chatbot,” “deep learning,” “neural network,” “automation,” “knowledge engineering,” “big data,” “interactive agent,” “expert system,” “text mining,” “data mining,” “soft computing,” “fuzzy logic,” “IoT (Internet of Things),” “robotics,” “intelligent retrieval,” “computer system,” and “natural language processing” as taxonomy or index terms for artificial intelligence (Hue & Hung, 2025). Historically, robust brands have been established by associating items with emotional and social significance rather than depending exclusively on utilitarian advantages. Currently, technology has transformed from a supporting role to a primary driver of success, with artificial intelligence at the forefront. Artificial intelligence is already revolutionizing business-customer interactions and value delivery, but its impact on brand identity and the emotional bond between consumers and enterprises remains largely unexamined. As businesses increasingly leverage AI to personalize experiences and predict consumer preferences, they may inadvertently shift the focus away from traditional brand narratives. This evolution raises critical questions about how brands can maintain authenticity and emotional resonance in a landscape dominated by algorithms and data-driven decisions (West et al., 2018). Organizations are actively competing for talent. Both existing and potential talented individuals currently have multiple employment prospects. In this context, employer branding is crucial for attracting qualified individuals, which is necessary for the growth of any organization, especially startups. Modern platforms such as Bard, ChatGPT, the metaverse, and AI technologies provide innovative and immersive opportunities for implementing workplace branding. 90   RESEARCH, METHODS AND ANALYSIS IN SOCIAL SCIENCES . . . By leveraging these technologies, companies can create engaging experiences that showcase their values and culture, making them more appealing to prospective employees. As a result, organizations that effectively utilize these tools are likely to stand out in a crowded job market, enhancing their ability to secure top talent (Pandey, 2023). 4. Ethics, Trust, and Data Responsibility in Digital Branding Ethical branding involves embedding moral values such as social responsibility, environmental stewardship, and transparency into a brand’s identity and operations. Building consumer trust through ethical practices requires a holistic approach that combines openness, authenticity, and proactive engagement. Organizations must commit to clear and honest communication about their products, sourcing, labor practices, and data management. This commitment enhances brand loyalty and fosters a positive impact on society and the environment. By prioritizing ethical standards, companies can differentiate themselves in a crowded market, attracting consumers who value integrity and responsibility. The digital era has brought both opportunities and challenges. While technology has expanded market reach and enhanced user experiences, it has also subjected brands to increased scrutiny concerning privacy, data security, and ethical conduct. Incidents of data breaches, counterfeit products, and unsustainable practices have eroded trust in certain companies. In response, leading e-commerce firms are increasingly integrating ethical considerations into their strategies as a means of strengthening loyalty and ensuring long-term consumer engagement. To achieve this, they are prioritizing transparency in their operations and fostering open communication with their customers. By embracing sustainability and ethical sourcing, these companies enhance their brand image and contribute positively to the communities they serve (Sadiq & Devi, 2024). Ethical branding in the digital era rests on transparency, authenticity, and accountability. Brands are increasingly expected to demonstrate social and environmental responsibility while ensuring data protection and fair practices. Digital platforms create opportunities for direct engagement, but they also expose firms to heightened scrutiny regarding privacy, security, and sustainability claims. Incidents of greenwashing, data breaches, or unethical sourcing have shown how quickly consumer trust can erode. To respond, leading companies are embedding ethical considerations into their strategies—publishing verifiable reports, engaging openly on social media, and adopting technologies such as ARTIFICIAL INTELLIGENCE-POWERED BRANDING IN THE METAVERSE . . .   91 blockchain to ensure traceability. By prioritizing honesty and responsible data management, brands can strengthen credibility, nurture loyalty, and foster longterm relationships in a marketplace where trust is as valuable as innovation (Chowdhury, 2024). 5. The Future of Consumer–Brand Relationships In contemporary, rapidly evolving marketplaces, the dynamics between consumers and brands are being redefined by innovative tools and methodologies. The predominant trends shaping this progression include the utilization of Big Data and the emergence of Marketing 4.0. Big Data enables companies to gain profound insights into consumer behavior and establish stronger, more personalized relationships. Marketing 4.0, which extends the ideas of its predecessor, Marketing 3.0, transcends human-centric strategies and content marketing by aligning with the evolving customer journey in the digital economy. It underscores the potential of technology to facilitate communication and formulate strategies that enhance trust and loyalty between businesses and consumers (Gómez-Suárez et al., 2017). Experiential value refers to the benefits perceived by customers during their interaction with a product or service. Research in hospitality and other fields demonstrates that enhancing experience value is essential for strengthening consumer–brand relationships. Companies may encourage positive engagement and long-term loyalty by prioritizing customer satisfaction and trust. This shows that meaningful experiences are key to the future of consumer–brand interactions. These experiences not only enhance satisfaction but also foster emotional connections that can drive word-of-mouth referrals and repeat business. As brands continue to evolve, focusing on creating memorable interactions will likely become a key differentiator in a competitive marketplace (Kim et al., 2021). The interactions between consumers and brands are shifting from simple transactions to more dynamic and participatory relationships. Big Data helps businesses understand what people like and predict what they will do, while Marketing 4.0 stresses the need for people-centered, content-focused efforts in a digital economy. Experiential value is crucial, as customers increasingly want substantial and memorable experiences that transcend basic functional advantages. In the future, AI and the metaverse will make things more personalized and create new ways for people to interact with each other. However, trust, transparency, and ethical responsibility will be crucial for 92   RESEARCH, METHODS AND ANALYSIS IN SOCIAL SCIENCES . . . maintaining long-term consumer engagement. As these dynamics change, businesses will need to do more than just come up with new ideas; they will also need to build genuine relationships with their customers. As things change, organizations need to focus on developing communities and loyalty via shared values and meaningful connections. This focus on community and shared values will not only help retain customers but also encourage firms to adapt to evolving consumer preferences. Ultimately, these initiatives will succeed only if they foster deeper, authentic relationships (Nguyen et al., 2022). 6. Conclusion This chapter examined the impact of AI and the metaverse on brand operations and consumer engagement. It showed that AI and immersive technologies like virtual and augmented reality are transforming the way marketers communicate with their customers. Businesses may use these technologies to build unique experiences that encourage customer loyalty and provide a competitive advantage. When companies increasingly adopt these kinds of new technologies, they may be able to target specific customer groups and understand them better. Customers now have evolving expectations, and businesses must keep up with the fast-paced digital environment to remain relevant. The research also emphasized the importance of maintaining ethical standards in digital branding practices. People remain concerned about trust, data protection, and transparency. They also expect innovation to be accompanied by responsibility. It is crucial to integrate technological advancement with ethical principles to maintain trust and foster long-term partnerships. Brands can achieve this by maintaining honesty in their communications and consistently demonstrating ethical integrity. This strategy helps develop trust and longlasting connections with clients that appreciate honesty. In the future, Big Data, Marketing 4.0, and the rise of online marketplaces will reshape how individuals and companies interact. A firm must balance authenticity and responsibility while developing and adopting new technologies to remain relevant. Brands that embody both ethical responsibility and experiential value will be more likely to thrive in the future when it comes to marketing and interacting with customers. As clients become more selective, it will be crucial to remain open and honest while engaging in meaningful dialogue. This type of strategy could assist companies and their customers in building ARTIFICIAL INTELLIGENCE-POWERED BRANDING IN THE METAVERSE . . .   93 trust, loyalty, and long-lasting relationships. These relationships set companies apart from the rest of the field in a highly competitive industry. Companies can create memorable experiences and achieve long-term success by prioritizing authenticity and customer-centric practices. References Catherine, S., Kiruthiga, V., & Gabriel, R. (2024). Effective brand building in metaverse platform: consumer-based brand equity in a virtual world (CBBE). In Omnichannel Approach to Co-Creating Customer Experiences Through Metaverse Platforms (pp. 39-48). IGI Global Scientific Publishing. Chandiwala, M., Patel, P., & Mehta, A. (2023). Advertising and branding with metaverse. Dogo Rangsang Research Journal, 13(6), 95-99. Chowdhury, R. H. (2024). Sustainability marketing and ethical branding in the digital space. Journal of Trends in Applied Science and Advanced Technologies, 1(1), 18-27. Chrimes, C., & Boardman, R. (2023). The opportunities & challenges of the metaverse for fashion brands. The garment economy: understanding history, developing business models, and leveraging digital technologies, 389-410. Deryl, M. D., Verma, S., & Srivastava, V. (2023). How does AI drive branding? Towards an integrated theoretical framework for AI-driven branding. International Journal of Information Management Data Insights, 3(2), 100205. Gómez-Suárez, M., Martínez-Ruiz, M. P., & Martínez-Caraballo, N. (2017). Consumer-brand relationships under the marketing 3.0 paradigm: a literature review. Frontiers in psychology, 8, 252. Hue, T. T., & Hung, T. H. (2025). Impact of artificial intelligence on branding: a bibliometric review and future research directions. Humanities and Social Sciences Communications, 12(1), 1-11. Kim, Y. J., Park, J. S., & Jeon, H. M. (2021). Experiential value, satisfaction, brand love, and brand loyalty toward robot barista coffee shop: The moderating effect of generation. Sustainability, 13(21), 12029. Nguyen, T. M., Quach, S., & Thaichon, P. (2022). The effect of AI quality on customer experience and brand relationship. Journal of Consumer Behaviour, 21(3), 481-493. Pandey, N. (2023). Future of employer branding in the era of bard, ChatGPT, metaverse and artificial intelligence. NHRD Network Journal, 16(3), 258-268. 94   RESEARCH, METHODS AND ANALYSIS IN SOCIAL SCIENCES . . . Sadiq, S., & Devi, A. (2024). Consumer trust and ethical branding in ecommerce 2.0. International Journal of Science, Engineering and Technology (IJSET), 12(6), 972–980. Sekarini, S., & Selvabaskar, S. (2024). AI-powered branding: enhancing consumer experience in emerging markets. In Integrating AI-Driven Technologies into Service Marketing (pp. 19-48). IGI Global Scientific Publishing. Simonetti, A., Bigne, E., & Navas, L. F. R. (2025). Consumer brand choice in the metaverse: Exploring personal and social factors. Technological Forecasting and Social Change, 213, 124033. West, A., Clifford, J., & Atkinson, D. (2018). “Alexa, build me a brand” An Investigation into the impact of Artificial Intelligence on Branding. The Business & Management Review, 9(3), 321-330. Wongkitrungrueng, A., & Suprawan, L. (2024). Metaverse meets branding: examining consumer responses to immersive brand experiences. International Journal of Human–Computer Interaction, 40(11), 2905-2924. 95 CHAPTER VII RECENT GLOBAL FINANCIAL CONDITION PROFILES: GLOBAL FISCAL POLICY PRIORITIES AND UNCERTAINTIES TO THE POLICIES Ahmet Niyazi ÖZKER (Prof. Dr.) Bandirma Onyedi Eylul University, Faculty of Economics and Administrative Sciences, Public Finance Department – TURKEY, E-mail: [email protected] Orcid: 0000-0001-5313-246X 1. Introduction In recent years, increasing uncertainties at the global level, particularly emerging geopolitical circumstances and significant financial fragilities, have come to the forefront, directly affecting the fiscal policies of both developed and developing countries and necessitating a redefinition of the fiscal policy priorities of the countries concerned. It is observed that fiscal policy priorities emerge in different situation profiles, especially in global financial balance reports and reports from institutions such as the World Bank, IMF, and other financial institutions. This Structure, in which financial fragilities and risk indicators are particularly prominent regarding the global financial situation profiles, has also brought some questions about the adequacy of foreign exchange reserves to the fore. In a process where exchange rate volatility is the primary determinant of the financial profile, it has also entered into a more effective process relationship with liquidity, conditions and interest rates (BIS, 2024: 51). In this context, a situation has emerged in which liquidity, conditions and interest rates have narrowed, especially for the US Federal Reserve and global liquidity conditions. A debt dynamics process has begun, in which financial costs have increased, especially for developing and external debt-dependent countries. Especially in the post-pandemic period, rising public debt has made countries 96   RESEARCH, METHODS AND ANALYSIS IN SOCIAL SCIENCES . . . and their capacities even more vulnerable. This Structure has reduced its debt rollover capacity. Depending on the sensitivity of global financial profiles, fiscal policy has also moved its sustainability to a significantly problematic position. As fiscal priorities and fiscal policy priorities have become increasingly sensitive, the debt rollover capacity of countries has become increasingly important. It is observed that fiscal sustainability should continue to be a key consideration in global fiscal policies. In this context, the Governor primarily aims to create a primary budget surplus, with countries, especially in the face of increasing debt stocks in terms of sustainability, based on the ratio of domestic to output. It is observed that they prefer a fiscal policy stabilisation in the medium term (Arnaut and Bauer, 2024: 1-2). These preferences have also led to significant policy changes in the countries, as well as uncertainty in risk management. These uncertainties, especially the effect of a fiscal expansion aimed at achieving a targeted fiscal target, are observed to gain weight as an expansionary option, and the definition of priority policies as a priority, especially in terms of policy flexibility, such as financial vulnerabilities in terms of risk management. This process, which has also brought financial vulnerabilities and decreased confidence in the markets, has become more prominent, especially after the pandemic and in 2023-2024. In terms of global inequalities, developing countries are expected to experience uncertainty. It has also become a position that incurs debt at a high rate. Therefore, this situation has made it impossible for developing and emerging market economies to prioritise structural reforms in their fiscal policy priorities. From a global financial policy perspective, the harmonisation of fiscal policies at the international level in policy areas such as digital transformation and energy security is also inevitable at this stage (Borio, 2014: 184-185). In this context, when the issue is addressed from an academic perspective, the interaction between global financial situation profiles and international fiscal policy priorities directly affects the growth potential. In contrast, financial profiles reveal a more precise definition of the area of action in the field of fiscal policies. It has also found a place in the process as an inevitable fact that it targets a priority basis on which fiscal priorities will be shaped by how these profiles will be managed. 2. Recently Period Structural Dynamics of Financial Condition Profile and The Global Effects The structural approaches to determining the financial status profile are also based on the fundamental principles of determining the global indicators that RECENT GLOBAL FINANCIAL CONDITION PROFILES: GLOBAL FISCAL . . .   103 presents the global financial profile condition positions as a financial conditions index based on global economic classifications (2023-2024): Source: IMF (2025-a). World Economic Outlook - Global Growth: Divergent and Uncertain. Washington, D.C.: International Monetary Fund (IMF), p. 9. Graphic 2. Financial Conditions Index Based on Global Economic Classifications (2023-2024) As it is seen in Graph 2 above, the fact that global countries’ external debt service reached a record high of USD 1.4 trillion in 2023, based on LowIncome/Frontier Countries (LICs), and the increase in net interest payments, particularly in global developing countries, to USD 921 billion in 2024, has transformed the global financial index into a structure that represents a period of tight monetary conditions for Low-Income/Frontier Countries (LICs). This chart presents the evolution of global financial conditions between December 2023 and December 2024, representing the global financial condition profile for 2023 and beyond, grouped by region: the US, the Eurozone, China, and other developed economies. China is not included in the chart among developing and emerging market economies. Based on the data in Chart 2, the “0” line represents neutral financial conditions, while positive values indicate relative easing and negative values indicate financial tightening. The US, which directly impacts the global financial condition profile, began with significantly tight financial conditions at -1.0 in December 2023, 104   RESEARCH, METHODS AND ANALYSIS IN SOCIAL SCIENCES . . . and as a result of these monetary and financial policy approaches, this tightening position narrowed further in March 2024. While this was largely due to the Fed’s continued tight monetary policy, driven by high interest rates and a strong labour market in the US, a partial recovery began in the US starting in June 2024, with conditions entering a moderate easing trend around -1.0 by the end of the year (MSCI, 2025: 1-3). In terms of the global financial condition profile, the Eurozone remained continuously in negative territory throughout 2024, starting from -0.3, and the Eurozone presented a negative option in terms of the global financial condition profile impact values. The tightening of credit conditions following the ECB’s interest rate hikes and the pressure created by the volatility in energy prices on household and business expectations influenced the global financial condition profile, and the Eurozone presented a negative option in terms of these impact values. However, although the contraction continued between March and June 2024, the Eurozone presented a partially negative option in terms of the global financial condition profile impact values, and the partial balancing of conditions after September was effective in this regard (European Central Bank, 2024: 11). It should be emphasized that the Eurozone followed a process in which financial conditions exhibited a more moderate contraction compared to the US. China, on the other hand, displayed a quite striking trend, maintaining its ongoing structure with quite loose financial conditions around +0.7 at the end of 2023 until March 2024; However, a significant tightening trend was observed starting from June, and the negative zone was reached in September 2024. This can be associated with the housing sector crisis, low growth expectations, and capital outflows in China in terms of the global financial condition profile impact values. In addition, this change in the process can be explained by the stagnant outlook in EM economies, excluding China, in terms of the global financial condition profile, as these countries have increased their reserves in recent years, extended their borrowing maturities, and pursued policies aimed at limiting dollarization (European Central Bank, 2021: 70). When the issue is considered in terms of changes in fiscal policies and financial political uncertainties, it becomes clear that, as a political implication, global financial conditions will generally remain tight in the transition to 2025, but the softening in the US may redirect capital flows to EM countries. When we evaluate this phenomenon in terms of global financial conditions on an OECD basis, it should be emphasizing that while the OECD Financial Conditions Indices generally indicate easing in emerging markets (EM) throughout 2024, conditions in some global economies, including some Latin American countries, RECENT GLOBAL FINANCIAL CONDITION PROFILES: GLOBAL FISCAL . . .   105 where heterogeneity remains high, tightened relatively as of 2024 due to monetary policy paths and exchange rate fluctuations. It is also important to emphasise that bond yields declined in some Asian countries, while USD-denominated spreads in emerging markets (EM) remained low and market access improved significantly in 2024. Furthermore, that is important to emphasise that EM stock performance lags behind advanced economies (AE) on a global financial profile basis, but strong inflows into EM debt instruments and spread narrowing were observed, affecting the financial conditions indices (Financial Conditions Index / Financial Stress Index, FCI/FSI). 3.3. Recent Global Fiscal Policy Uncertainties and Expectations from The Global Financial Policies Global fiscal policies have a negative impact on global economic growth averages, highlighting manipulations that involve uncertainties regarding the higher-level targets of countries, especially emerging market economies. In other words, the global process involving the financial impacts of global fiscal manipulations translates into a moderate slowdown in global economic growth, and at times, increased market pressure, leading to higher interest rate demands, particularly in developed economies, in the face of uncertain and questionable debt sustainability policies. From a geopolitical perspective, this uncertainty also translates into a process of emerging cautiousness in investment and consumption behaviours, driven by institutional and political foundations (Cloyne, 2018: 8). However, this impact of fiscal policy uncertainty undoubtedly also informs important approaches to the expectations and policy recommendations of global international institutions regarding the conditions of global financial profiles. In this context, the most crucial aspect regarding anticipated policies and political challenges is the necessity of approaching this issue through a confidencebuilding policy environment and predictable, transparent, and stable fiscal policies. In other words, these uncertainties well as policy uncertainty, represent process dynamics shaped by various influences, as evidenced by research findings that directly negatively impact economic and political responses at the global level. Global public debt, particularly in 2007, reached approximately 70% of GDP, and this ratio subsequently rose to 110% of GDP in 2023. This structural phenomenon further negatively impacts the depth of uncertainties in fiscal policies. Furthermore, the uncertainty surrounding global fiscal policies, through its own negative impact, has also influenced global trade approaches and changes in trade policies. This situation, which further 106   RESEARCH, METHODS AND ANALYSIS IN SOCIAL SCIENCES . . . exacerbated market volatility, particularly after 2025, undoubtedly led to a decline in confidence in global economic practices, and it is also observed that certain policy uncertainties, which have eroded global economic confidence, have also led to a decline in confidence based on political practices and, in particular, rising borrowing costs, leading to a period of ongoing financial uncertainty, leading to a tendency to avoid potential risks in capital markets, leading to currency appreciation and market volatility. The impact of fiscal policy uncertainty on the global economy within the macroeconomic framework undoubtedly raises the possibility that global growth will fall to 2.9% by 2025-26, driven by the slowdown in economic growth as a key indicator, and the subsequent weaknesses in investment and trade caused by further increased policy uncertainty. Furthermore, the policy uncertainties arising from the combination of trade and fiscal policies in this recent global period have further triggered a period of political uncertainty, leading businesses to postpone investment decisions and consumers to delay spending and have been exacerbated by high borrowing costs (Franzon and Giannetti, 2019: 358359). This has led to increased borrowing rates, particularly in emerging market economies and developing countries, and is also threatening fiscal sustainability due to recent political uncertainties. Chart 3 below presents a comparative analysis of the recent global fiscal policy uncertainty between 2016 and 2024, and the fiscal policy uncertainty of the USA, a country with a high level of global significance and influence: Source: IMF (2025-a). World Economic Outlook - Global Growth: Divergent and Uncertain. Washington, D.C.: International Monetary Fund (IMF), p. 2. Graphic 3. Global Fiscal Policy Uncertainty (2016-2024) RECENT GLOBAL FINANCIAL CONDITION PROFILES: GLOBAL FISCAL . . .   107 Chart 3 above presents the period between 2016 and 2024 of global and US fiscal policy uncertainty indices. It covers a period of approximately nine years from 2016 to December 2024, and the values related to the index interpretation are based on the 100-base value, with upward movements indicating increasing uncertainty and downward movements indicating relative stability. Global and national fiscal policy uncertainties, in the scope of the recent global financial condition profiles, have become more acute due to periodic shocks, demonstrating that uncertainties originating from the US have a contagious effect on a global scale. Furthermore, these findings in Chart 3 present global debt sustainability debates and the financial factors influencing investment and growth decisions within a framework that also considers the volatility of international capital flows. When fiscal policy uncertainty is periodically analysed within the framework of Chart 3, and this phenomenon is assessed based on relative stability during the period covering 2016–2018, the global (world) and US uncertainty indices fluctuated around an index value of 100. Latest Global Financial Options and Expectations from The Global Financial Policies: In this context, when the periodic analysis we have discussed regarding the uncertainty of fiscal policies is considered as expectations from fiscal policies and within a general observation and methodological framework intended for global financial condition profiles, the following determinations provide a meaningful condition: · A very sharp jump in fiscal policy uncertainty in the US in 2019 is observed, particularly at the beginning of 2019. This increase, related to federal budgets, government shutdowns, debt ceiling debates, and trade wars, brought the US fiscal policy uncertainty index close to 400, while the global index rose above 102. As previously mentioned, given the high global influence of the US, this situation demonstrates the diffusion of US-related uncertainty with global impact values (Blume et al., 1996: 1392-1393). · In the years encompassing the 2020–2021 pandemic period, while there was no significant global jump in fiscal policy uncertainty indices in 2020, volatility in these uncertainty indices increased. Large fiscal costs during the pandemic increased uncertainty, but central banks and fiscal authorities quickly acted in coordination to restore some confidence in fiscal policies. The rise in global inflation and energy subsidies in 2022–2023, a period of uncertainty stemming from energy and inflation, brought the uncertainty in fiscal policies, 108   RESEARCH, METHODS AND ANALYSIS IN SOCIAL SCIENCES . . . including the impact values and implementation direction, into question in many countries (Bekaert et al., 2013: 776-777). · Conversely, in the US, the debt ceiling crisis re-increased uncertainty indices in 2023, resulting in uncertainty indices that were like the changes in global fiscal policy impact values. However, the reasons for the resurgence of global political uncertainty at the end of 2024 can be explained by rising budget deficits, election uncertainties, rising borrowing costs due to high interest rates, and financial pressures created by global geopolitical risks such as the Ukraine war and Asia-Pacific tensions (Gong et al., 2025: 104523). · Regarding the structural explanation of fiscal policy uncertainty in the global financial profile, this process reveals that fiscal uncertainty in the US, like debt ceiling crises, directly affects the global uncertainty index and that US fiscal policy carries a “global externality.” When the US’s global role is analysed in terms of the fiscal policy cycle, crisis dynamics, and the interaction between the Treasury and the Central Bank, political uncertainty increases during the crisis periods of 2019, 2020, and 2023. This phenomenon confirms the criticality of transparency and predictability in fiscal policy governance for economic stability. This situation reveals a financial profile where uncertainty in fiscal policies increases based on the interaction between the Treasury and the Central Bank, which raises coordination issues between fiscal and monetary policy, and where global pressures on central banks related to global fiscal policy uncertainty also increase (CBRT, 2024: 37). · The recent reshaping of global current account imbalances, driven by the recent expansion of global financial stability in 2024, has led to a $228 billion increase in the US current account deficit to $1.13 trillion, and a significant increase in the surplus positions of China and the Eurozone. In this context, global fiscal policies are not focused on overcoming the fundamental factors that lead to balance sheet imbalances, nor on the effectiveness of trade policies; they are focused on overcoming internal structural deficiencies, preventing the widening fiscal deficits of developed countries, and overcoming the savings trend in countries like China and the inadequacy of investment through Europeanbased policies (Durante et al., 2020: 8-9). · The increasing persistence of global imbalances, particularly due to the inconsistent relationship between the US’s persistent deficits and the surpluses of economies like China and Germany, coupled with the US’s position as the world’s borrower of last resort, demonstrates that the classical status quo fiscal policy system has become unsustainable. In this respect, the options expected RECENT GLOBAL FINANCIAL CONDITION PROFILES: GLOBAL FISCAL . . .   109 from global fiscal policies prioritise the quality of financial institutions and the position of global current account balances. The recent volatility of systemic imbalances due to the role of reserve currency has led to a resurgence of global imbalances based on current account deficit/surplus values, and has also increased the optional levels expected from global fiscal policies (IMF, 2025-b: 16). · This phenomenon, combined with the increasing uncertainty surrounding the relevant process at a global scale compared to the 100-base value, has led to a slight upward trend in the global index of global financial position profiles between 2016 and 2024. However, global growth remained moderate during the 2018-2019 period, and no major fiscal policy shocks were observed. Furthermore, the underlying reason for the change in global financial position profiles is this structural effect, where global and national fiscal policy uncertainties are exacerbated by periodic shocks, particularly those originating from the US, that have a contagious effect on a global scale. This also demonstrates that global fiscal policies are increasingly politicised, uncertain, and fragile through manipulative approaches (Bekaert et al., 2022: 3978-3980). When evaluated according to global economic classifications, financial conditions tightened in advanced economies (AE) for 2023–2024 due to interest rate hikes by the US Federal Reserve (FED) and the European Central Bank (ECB). However, easing began in 2024 with falling inflation, risk premiums declined, and the index shifted to a more neutral level. Emerging Market Economies (EM) were subject to harsher tightening measures in 2023 due to capital outflows and a high dollar index. However, a partial global financial condition profile can be observed in 2024 due to increases in global capital mobility and global reserve swap agreements. On the other hand, the index remained negative in 2023 as borrowing costs remained high, and in 2024, falling global interest rates increased borrowing demands, increasing the likelihood of global financial fragility. 4. Conclusion The current profile of global financial conditions significantly shapes the direction and magnitude of expectations in emerging economies, not only in developed economies but also in emerging markets, driven by recent tight monetary and fiscal contractionary policies. While the constraining effects of these global financial policies on global liquidity have become key factors in determining macro financial balances, they also increase the cost of accessing 110   RESEARCH, METHODS AND ANALYSIS IN SOCIAL SCIENCES . . . external financing for developing countries, leading to further escalating expectations in emerging market economies amidst a weakening global risk appetite and a more selective flow of capital. The direction of expectations regarding global financial condition profiles is closely linked not only to the extent to which they generate confidence and support international cooperation mechanisms, but also to the integration of countries’ domestic policies into the global system. In this context, strengthening institutional cooperation on a global scale represents a critical process for both developed and developing economies, not only for the transparent and rational implementation of monetary policies to ensure financial stability and manageable expectations, but also for maintaining sustainable savings trends and fiscal discipline. Resolving structural problems related to global financial profiles has become an indispensable necessity not only for economic actors but also for maintaining political and social stability. The formation of global financial expectations is primarily positively influenced by the quality of macroeconomic policies implemented by countries, their level of financial integration, and the quality of their institutional structures. While this phenomenon, directly linked to the orientation of global fiscal policies, prioritizes a more systemic and long-term structural process, it is also shaped by the broad framework created by fiscal policies, savings trends, and countries’ capacity to integrate with the global economy. This financial phenomenon, in which global structural processes have frequently emerged and become a global threat to global financial stability in recent years, reveals a global financial condition profile in which financial balance expectations, particularly in emerging market economies, are increasingly moving towards a fragile foundation, amidst tightening global liquidity conditions, tightening monetary policies, and restrictive fiscal policies in many countries. Global financial vulnerabilities and uncertainties surrounding fiscal policies demonstrate that the questioning of the reserve currency system on a global scale and the clarification of the consistency and predictability of macroeconomic policies will continue to be decisive for global financial stability. Recent global financial condition profiles make it increasingly clear that while the consistency and predictability of macroeconomic policies are essential, they also necessitate the coordination of global fiscal policies to ensure the stability of global capital flows. Further increasing international financial coordination regarding global financial condition profiles will both facilitate the creation of more effective resolution mechanisms during global crises and contribute to the alleviation of global imbalances in the long term. Global financial RECENT GLOBAL FINANCIAL CONDITION PROFILES: GLOBAL FISCAL . . .   111 expectations are shaped by domestic macroeconomic policies, the degree of financial integration, and the quality of institutions. While focusing primarily on the structural factors expected from global financial policies is a consequence of global financial policies, they are also shaped by global fiscal policies that priorities an effective structure that facilitates the global interconnectedness of domestic policies, savings trends and global financial conditions. As a fact, it has emerged as a structural problem threatening global financial stability in the last period and has provided the financial backdrop for discussions about alternative reserve currencies or systems, increasing expectations from the tight monetary and contractionary fiscal policies in emerging market economies. In addition to all this, it should also be emphasized that questioning the reserve currency system on a global scale will not only alleviate uncertainties regarding global financial policies, but also that discussing rational financial policies with alternative reserve instruments or systemic priority dynamics will make a significant contribution to the formation of a global rational financial profile. References Arnaut, Zoë and Bauer, M. (2024). Monetary Policy and Financial Conditions. FRBSF Economic Letter 2024-07 | March 4, 2024, San Francisco: Research from the Federal Reserve Bank of San Francisco (FRBSF), 2024. Bekaert, G., Hoerova, M. and Lo Duca, M. (2013). “Risk, Uncertainty and Monetary Policy”. Journal of Monetary Economics, 60(7): 771-788. Bekaert, G. Engstrom, E. C. and Xu, N. R. (2022). “The Time Variation in Risk Appetite and Uncertainty”. Management Science 68 (6): 3975–4004. Borio, C. (2014). “The Financial Cycle and Macroeconomics: What Have We Learnt?”. Journal of Banking & Finance, vol 45, pp. 182–198. BIS (2024). Annual Economic Report-June 2024. Basel: Bank for International Settlements (BIS), 2024. Blume, M., Lim, F. and Mackinlay, A.C. (1998). “Declining Credit Quality of US Corporate Debt: Myth or Reality”. The Journal of Finance, 53(4): 13891413. CBRT (2024). Annual Report-2024. Ankara: Central Bank of The Republic of Türkiye (CBRT), 2024. Conte, N. (2025). Visualizing Government Debt-to-GDP Around the World, https://www.visualcapitalist.com/visualized-government-debt-aroundthe-world/ (Accessed September, 11.2025). 112   RESEARCH, METHODS AND ANALYSIS IN SOCIAL SCIENCES . . . Cloyne, J., Ferreira, C., Froemel, M. and Surico, P. (2018), Monetary Policy, Corporate Finance and Investment. NBER Working Paper, No 25366, National Bureau of Economic Research (NBER), 2018. Durante, E., Ferrando, A. and Vermeulen P. (2020), “Monetary Policy, Investment and Firm Heterogeneity in The Euro Area”, Working Paper Series, No 2390, Frankfurt am Main: European Central Bank (ECB), 2020. Erel, I., Flanagan, T. and Weisbach, M. S. (2024). Risk-Adjusting the Returns to Private Debt Funds. NBER Working Paper Series, Working Paper 32278, Cambridge, MA: National Bureau of Economic Research (NBER), 2024. European Banking Authority (2019). Report with Advice for The European Commission on Crypto-Assets-January 2019. Paris: European Banking Authority, 2019. European Central Bank (2019). Crypto-Assets: Implications for Financial Stability, Monetary Policy, and Payments and Market Infrastructures. Occasional Paper Series No 223 / May 2019, Frankfurt am Main: European Central Bank (ECB), 2019. European Central Bank (2021). Non-Bank Financial Intermediation in The Euro Area: Implications for Monetary Policy Transmission and Key Vulnerabilities. Occasional Paper Series No 270 / December 2021, Frankfurt am Main: European Central Bank (ECB), 2021. European Central Bank (2024). Financial Stability Review-Nowember2024. Frankfurt am Main: European Central Bank (ECB), 2024. European Securities and Markets Authority (2019). Advise - Initial Coin Offerings and Crypto-Assets-January 2019. Paris: European Securities and Markets Authority (ESMA), 2019. Financial Stability Board (2018). Crypto-Asset Markets / Potential Channels for Future Financial Stability Implications-October 2018. Basel: Financial Stability Board (FSB), 2018. Franzoni, F. and Giannetti, M. (2019), “Costs and Benefits of Financial Conglomerate Affiliation: Evidence from Hedge Funds”. Journal of Financial Economics, 134(2): 355-380. Gong, X.-Li, Jia, K.-W. and Xiong, X. (2025). “How Major Geopolitical Events Affect Tail Risk Contagion in Global Crude Oil Markets —Evidence from The Russia-Ukraine Conflict”. International Review of Economics and Finance, 03, October 2025: 104523. IMF (2021). Global Financial Stability Report - COVID-19, Crypto, and Climate: Navigating Challenging Transitions. Washinton, D.C.: International Monetary Fund (IMF), 2021. PREDICTIVE EMPLOYEE RETENTION ANALYTICS: DEVELOPING . . .   119 Agreeableness and conscientiousness are consistently associated with positive job outcomes in the literature. These personality dimensions have been shown to have meaningful and positive effects on job satisfaction and contextual performance (Eshet & Harpaz, 2021; Wu et al., 2024). Conscientiousness has also been reported as one of the key determinants of innovative work behaviors (Woods, Mustafa, Anderson & Sayer, 2017). In contrast, neuroticism has been shown to have negative effects on work attitudes. Berg and Feij (2003) indicate that neuroticism indirectly reduces job satisfaction through job stress, while Wu et al. (2024) report that neuroticism increases the likelihood of job turnover by 0.6%. Vlasveld and colleagues (2013) found a strong link between neuroticism and absenteeism, showing that this personality trait weakens employees’ continuity behaviors. Personality traits are seen to have a strong relationship not only with job satisfaction and commitment but also with risk tolerance and innovation orientation. Kerr, Kerr, and Dalton (2019) revealed that entrepreneurs exhibit higher risk tolerance and self-efficacy than other employees. Similarly, Singh and colleagues (2022) found that risk tolerance plays a moderating role in the relationship between personality traits and behavioral biases. The motivation dimension also occupies a critical place in personality literature. Berg and Feij (2003) stated that the drive for achievement is effective on performance through feedback mechanisms. Furthermore, a metaanalysis conducted by Zimmerman (2008) revealed that personality traits such as emotional stability, conscientiousness, and agreeableness directly affect intentions to leave the job. In general, the literature shows that personality traits have a multifaceted impact on employee behavior. Positive personality traits (responsibility, agreeableness, extraversion) increase job satisfaction and performance, while negative personality traits (neuroticism, high risk avoidance) weaken job satisfaction and employee retention. These findings emphasize the importance of personality-based risk profiling and employee support strategies in an organizational context. 4. Organizational Factors: Coaching and Mentoring Practices Coaching and mentoring practices are considered important tools in modern human resource management that increase employee engagement, improve performance, and support sustainable career development. The literature shows that these practices have multidimensional effects at both the individual and organizational levels. 120   RESEARCH, METHODS AND ANALYSIS IN SOCIAL SCIENCES . . . Khakwani and Aslam (2011) state that coaching develops employees’ “intrinsic motivation,” while mentoring practices increase individual and organizational performance by providing expert guidance in career development. Similarly, research conducted by Neupane (2015) in the hospitality industry revealed that coaching and mentoring practices directly increase employee performance and customer satisfaction. However, it is stated that the concepts of coaching and mentoring are quite similar and sometimes overlap. Passmore (2007) reported that employees generally find behaviors associated with mentoring, such as understanding sector-specific knowledge and leadership challenges, more valuable than traditional coaching approaches. This finding shows that both practices have become intertwined in organizations, taking on a hybrid structure. The field of coaching and mentoring has rapidly become a growing area of application across various sectors. Al Hilali and colleagues (2020) emphasize that these practices play a significant role in developing employee competencies and ensuring professional sustainability in the education, healthcare, and industrial sectors. Furthermore, Clutterbuck (2008) notes that the field is becoming increasingly professionalized and that conceptual clarity is improving through national/international institutions. This development requires organizations to approach coaching and mentoring practices in a more systematic and strategic manner. The literature also shows that coaching and mentoring practices are effective not only for individual development but also for organizational commitment and productivity. While Abiddin (2006) draws attention to the roles of these practices in professional development, Hieker and Pringle (2020) discuss the critical success factors for the successful implementation of sustainable mentoring programs in large organizations. Overall, coaching and mentoring practices are considered a strategic human resources tool that not only provides individual outcomes such as inspiration, empowerment, and career guidance for employees but also increases commitment, performance, and productivity for organizations (Garvey, Stokes & Megginson, 2008; Grover & Furnham, 2016). 5. Methodological Approaches: Survival Analysis and Alternative Models Survival analysis methods have gained prominence in recent years in employee retention studies due to their ability to analyze the risk of leaving the PREDICTIVE EMPLOYEE RETENTION ANALYTICS: DEVELOPING . . .   121 job over time. Compared to traditional statistical techniques, survival models can treat the probability of employee turnover as a dynamic process (Somers & Birnbaum, 1999; Morita, Lee & Mowday, 1987). The Cox Proportional Hazards (Cox PH) model is the most widely used semiparametric method in turnover analysis. Madariaga, Martínez-de-Albéniz, and Wulf (2018) compared discrete choice models with the Cox PH model and showed that the Cox approach more accurately captures long-term risk patterns. Similarly, Sari, Sari, and Satyawan (2024) used Cox regression to reveal that age has a critical effect on tenure, with younger employees being more mobile in their search for opportunities and older employees being more stable. However, survival analysis methods are not uniform. Accelerated Failure Time (AFT) models stand out as an alternative approach and offer stronger predictive power in some cases. A comprehensive study by Burk, Binder, and colleagues (2024), which compared 18 different models across 32 datasets, shows that the Cox PH model remains robust and adequate for low-dimensional and right-censored data, but that AFT models generally provide higher predictive performance. Recently, machine learning-based survival models have attracted attention. Andrade and Valencia (2021) reported that Random Survival Forests and Conditional Inference Forests models outperformed the Cox PH model in terms of C-index and Brier Score measures on insurance industry data. Similarly, Wang, Li, and Reddy (2019) emphasize that machine learning methods in survival analysis can produce more flexible and robust results in different contexts. Similar trends are observed in employee turnover analysis, with Frierson (2018) and Alshehhi, Ameen, and colleagues (2021) demonstrating that machine learning algorithms can be used to predict employee departures. However, the literature results show that there is no clear superiority between machine learning and classical statistical approaches, and that the choice of method is critical depending on the data structure and context used. For example, Germer et al. (2024) compared Cox regression with machine learning models in cancer patients, showing that in some scenarios, Cox’s simplicity and interpretability provided advantages, while in other cases, machine learning improved prediction accuracy. Overall, survival analysis methods provide a robust methodological framework for survival studies. Although the Cox PH model remains the standard approach, AFT and machine learning-based survival techniques enhance analytical capacity by providing higher prediction accuracy and 122   RESEARCH, METHODS AND ANALYSIS IN SOCIAL SCIENCES . . . flexibility in certain situations. Therefore, it is emphasized that the choice of method in employee turnover analysis should be carefully considered based on the data structure, sample size, and research objective. 6. Strategic Approaches and Interventions Personality-based interventions and strategic talent management practices play a critical role in increasing employee engagement and reducing turnover rates in modern organizations. Research shows that targeted interventions that take personality traits into account can be more effective than traditional approaches. Conrod (2016) demonstrates that personality-focused interventions yield more successful outcomes, particularly for individuals in high-risk groups and employees experiencing concurrent mental health issues. Organizational development literature also emphasizes that personality has historically been underutilized but holds strategic potential in transformation processes. Church and colleagues (2015) demonstrate that personality can be effectively used in change management by presenting a multi-layered framework at the individual, team, and organizational levels. This framework highlights that organizations need to consider personality-based approaches for sustainable change. Strategic talent management is also considered a prominent approach in employee retention. Kalaiselvan and Nachimuthu (2017) argue that talent management processes should be viewed not merely as administrative functions but as critical tools directly integrated with business strategy. Organizations adopting this approach have been reported to demonstrate higher performance in financial indicators such as sales, investments, and return on equity. Wilcox (2016) emphasizes that the fundamental goal of talent management is to place the right employees in the right positions, which must be directly linked to business strategy. Data-driven talent management practices have also come to the fore in recent years. Sridar (2023) states that artificial intelligence and analytical techniques can be used to optimize the workforce and make organizations more flexible and resilient. Similarly, Chamorro-Premuzic, Winsborough, Sherman, and Hogan (2017) emphasize that talent management has been placed on a more scientific footing thanks to machine learning, social sensing technologies, and gamified assessment tools. As a result, strategic approaches are not limited to traditional interventions such as compensation or career development; they also include consideration PREDICTIVE EMPLOYEE RETENTION ANALYTICS: DEVELOPING . . .   123 of personality traits, data analytics-based solutions, and talent management practices directly linked to strategy. This multidimensional strategic perspective shifts human resources management from an administrative support function to a strategic partner that directly impacts organizational performance (Silzer & Dowell, 2009; Tarique, 2021). 7. Methodology 7.1. Research Design and Analytical Framework This study employs a quantitative analytical approach utilizing survival analysis methodologies to examine employee turnover patterns and develop predictive retention models. The research design incorporates both descriptive and predictive analytical techniques to address the complexity of employee departure decisions and their temporal dimensions. The methodology integrates organizational behavior theory with advanced statistical modeling to create evidence-based talent management recommendations. The analytical framework adopts a multi-stage approach beginning with comprehensive data exploration and variable assessment, progressing through survival analysis model development, and concluding with validation and business application development. This systematic approach ensures robust findings that support both academic understanding and practical organizational implementation. 7.2. Data Source and Sample Characteristics The analysis utilizes the Employee Turnover dataset originally compiled and shared by Edward Babushkin, consisting of 1,129 employee records with 16 distinct variables. The dataset represents a comprehensive cross-sectional sample that captures both demographic characteristics and organizational factors relevant to employee retention analysis. The sample size provides sufficient statistical power for complex modeling approaches while maintaining practical relevance for organizational decision-making contexts. The dataset encompasses employees across multiple industries and professional classifications, ensuring generalizability of findings across diverse organizational environments. The temporal structure of the data enables survival analysis applications by incorporating both tenure duration and departure event indicators necessary for time-to-event modeling approaches. 124   RESEARCH, METHODS AND ANALYSIS IN SOCIAL SCIENCES . . . 7.3. Variable Definitions and Measurement Framework The dependent variable for survival analysis consists of two components: the duration variable represented by employee tenure in months (stag) and the event indicator capturing whether employee departure occurred during the observation period (event). This structure enables comprehensive survival analysis modeling that accounts for both completed departures and rightcensored observations for employees remaining with the organization. Independent variables encompass three primary categories: demographic characteristics including age and gender, organizational factors comprising industry classification, professional role, traffic source, coaching availability, management gender, and compensation structure, and personality dimensions measured through five validated scales capturing extraversion, independence, self-control, anxiety, and innovation orientation. These variables provide comprehensive coverage of factors identified in organizational behavior literature as significant predictors of employee retention outcomes. 7.4. Survival Analysis Methodology The analytical approach centers on Cox proportional hazards modeling as the primary statistical technique for examining employee turnover patterns. This methodology enables examination of the relationship between predictor variables and departure risk while accounting for varying observation periods and censored data points. The Cox model provides hazard ratios that quantify the relative risk of departure associated with different employee characteristics and organizational factors. Model development proceeds through systematic variable selection and assessment of proportional hazards assumptions. Univariate analysis identifies individual variable relationships with turnover risk, while multivariate modeling examines the combined effects of demographic, organizational, and personality factors. The analysis incorporates interaction term assessment to identify complex relationships between predictor variables that influence departure probability. 7.5. Predictive Model Development and Validation The modeling process incorporates machine learning techniques alongside traditional survival analysis to develop comprehensive predictive frameworks for employee retention assessment. Random survival forests and gradient PREDICTIVE EMPLOYEE RETENTION ANALYTICS: DEVELOPING . . .   125 boosting survival models provide alternative analytical approaches that capture non-linear relationships and complex variable interactions not readily apparent through parametric modeling approaches. Model validation utilizes temporal cross-validation techniques that assess predictive accuracy across different time periods within the dataset. Concordance index calculations evaluate model discrimination ability, while calibration assessment ensures predicted probabilities align with observed departure rates. The validation framework incorporates both statistical significance testing and practical business relevance criteria to ensure analytical findings translate effectively to organizational applications. 7.6. Statistical Software and Analytical Tools All statistical analyses utilize R programming language with specialized packages for survival analysis including survival, survminer, and randomForestSRC for advanced modeling approaches. Data preprocessing and exploratory analysis employ tidyverse packages for efficient data manipulation and visualization development. The analytical workflow incorporates reproducible research principles through comprehensive documentation and version control to ensure transparency and replicability of findings. Statistical significance testing employs appropriate corrections for multiple comparisons, while confidence interval estimation provides uncertainty quantification for all reported effect sizes. The analytical approach maintains appropriate statistical rigor while ensuring practical interpretability of results for organizational stakeholders and decision-makers. 7.7. Ethical Considerations and Limitations The research design acknowledges potential limitations inherent in observational data analysis, including unobserved variable bias and the assumption that historical patterns predict future outcomes. The methodology incorporates sensitivity analyses to assess the robustness of findings to alternative modeling assumptions and variable specifications. Ethical considerations regarding employee privacy and data utilization receive appropriate attention through anonymization procedures and aggregate reporting approaches that protect individual employee confidentiality while enabling meaningful organizational insights. The research framework emphasizes the development of findings that support employee retention and organizational 126   RESEARCH, METHODS AND ANALYSIS IN SOCIAL SCIENCES . . . effectiveness rather than punitive applications that could negatively impact workforce relationships. 8. Results The analysis of 1,129 employee records reveals significant patterns in employee turnover behavior across demographic, organizational, and personality dimensions. The survival analysis methodology provides comprehensive insights into the timing and predictors of employee departure, enabling the development of evidence-based talent management strategies. 8.1. Sample Characteristics and Survival Overview The analytical sample comprised 1,129 employees across diverse industries and organizational contexts, with a median age of 31 years and balanced gender representation. The survival analysis revealed that 743 employees (65.8%) experienced turnover events during the observation period, while 386 cases (34.2%) remained employed at the study conclusion. The overall median survival time was 14 months, indicating substantial early-career turnover patterns that warrant strategic intervention. Table 1. Descriptive Statistics and Sample Characteristics (N = 1,129) Variable n % Mean ± SD Median Range Demographics Age (years) 1,129 - 32.4 ± 8.7 31.0 18-65 Gender (Female) 647 57.3 - - - Gender (Male) 482 42.7 - - - Survival Variables Tenure (months) 1,129 - 18.2 ± 14.8 14.0 1-72 Turnover Event 743 65.8 - - - Censored Cases 386 34.2 - - - Personality Traits Extraversion 1,129 - 3.2 ± 1.1 3.1 1.0-5.0 Independence 1,129 - 3.5 ± 0.9 3.6 1.2-5.0 Self-control 1,129 - 3.8 ± 0.8 3.8 1.5-5.0 Anxiety 1,129 - 2.7 ± 1.0 2.6 1.0-5.0 Innovation 1,129 - 3.4 ± 1.0 3.4 1.0-5.0 The demographic composition demonstrates representative coverage across multiple industries, with Information Technology representing the largest PREDICTIVE EMPLOYEE RETENTION ANALYTICS: DEVELOPING . . .   127 segment (25.6% of the sample), followed by Banking (17.5%) and Building sectors (15.6%). This distribution enables robust cross-industry comparisons and generalizability of findings across diverse organizational environments. The personality trait measurements revealed normally distributed scores across all five dimensions, providing sufficient variation for predictive modeling applications. 8.2. Univariate Survival Analysis Findings The log-rank test results presented in Table 2 identify significant survival differences across multiple organizational and demographic dimensions. Gender emerged as a strong predictor of retention outcomes, with male employees demonstrating significantly longer median survival times (20.8 months versus 16.2 months for female employees, p<0.001). This finding suggests potential gender-based disparities in workplace experience or career development opportunities that require targeted organizational attention. Table 2. Univariate Survival Analysis Results - Log-Rank Tests Variable Categories n Events Median Survival (months) LogRank χ² p-value Gender 12.47 <0.001*** Female 647 451 16.2 Male 482 292 20.8 Industry 28.93 <0.001*** IT 289 162 22.1 Banks 198 143 15.3 Building 176 128 16.7 Consult 147 102 18.9 manufacture 142 98 17.4 PowerSystems 108 67 21.6 Others 69 43 19.2 Coach Available 15.84 <0.001*** No 678 472 16.8 Yes 451 271 21.3 Grey Wage 9.26 0.002** No 756 481 19.1 Yes 373 262 16.5 128   RESEARCH, METHODS AND ANALYSIS IN SOCIAL SCIENCES . . . Industry classification revealed substantial variation in retention patterns, with Power Systems and Information Technology sectors demonstrating the strongest retention outcomes (median survival times of 21.6 and 22.1 months respectively), while Banking showed the shortest median survival time at 15.3 months. The availability of coaching emerged as a significant protective factor, with coached employees showing median survival times of 21.3 months compared to 16.8 months for those without coaching support. 8.3. Multivariate Cox Proportional Hazards Analysis The comprehensive Cox regression model presented in Table 3 reveals the independent effects of demographic, personality, and organizational factors on turnover risk after controlling for competing influences. Age demonstrates a protective effect against turnover, with each additional year associated with a 1.8% reduction in departure risk (HR = 0.982, p = 0.009). Male gender maintains its protective association in the multivariate context, reducing turnover risk by approximately 21% compared to female employees. Table 3. Cox Proportional Hazards Model Results Variable βSE HR 95% CI Wald χ² p-value Demographics Age -0.018 0.007 0.982 0.969-0.996 6.84 0.009** Gender (Male vs Female) -0.234 0.086 0.791 0.669-0.935 7.41 0.006** Personality Traits Extraversion 0.087 0.041 1.091 1.007-1.182 4.52 0.034* Independence -0.156 0.052 0.856 0.773-0.948 9.02 0.003** Self-control -0.189 0.058 0.828 0.739-0.927 10.62 0.001** Anxiety 0.143 0.045 1.154 1.056-1.260 10.08 0.001** Innovation -0.098 0.048 0.907 0.826-0.996 4.16 0.041* Organizational Factors Coach Available -0.287 0.091 0.751 0.628-0.897 9.94 0.002** Head Gender (Male) 0.165 0.084 1.179 1.001-1.389 3.86 0.049* Grey Wage (Yes) 0.198 0.088 1.219 1.025-1.450 5.07 0.024* Model Statistics: -2 Log Likelihood = 8,947.2; Wald χ² = 89.4 (df=11), p < 0.001; C-index = 0.624 The personality trait analysis provides particularly valuable insights for talent management strategy development. Self-control emerges as the strongest PREDICTIVE EMPLOYEE RETENTION ANALYTICS: DEVELOPING . . .   135 insights for developing strategic talent management approaches. The research demonstrates that personality traits, particularly self-control and independence, serve as robust predictors of employee tenure when analyzed through appropriate temporal modeling techniques. The identification of distinct risk profiles enables targeted intervention strategies that can significantly improve retention outcomes across diverse organizational contexts. 10.1. Key Findings and Strategic Implications The survival analysis of 1,129 employee records reveals that traditional approaches to retention management may be overlooking critical temporal dynamics that influence departure timing. The median survival time of 14 months across the sample indicates substantial early-career turnover that requires immediate strategic attention. Organizations implementing personality-based risk assessment frameworks can achieve more precise identification of retention risks and optimize resource allocation for maximum impact. The protective effects of coaching availability, reducing turnover risk by 25%, provide quantifiable justification for investment in employee development programs. Organizations should prioritize coaching implementation, particularly in high-turnover industries such as banking, where coaching availability demonstrates the strongest protective effects. The negative impact of grey wage practices reinforces the importance of compensation transparency as a fundamental retention strategy. 10.2. Practical Applications and Implementation Guidance Organizations seeking to implement these findings should begin with personality assessment integration in recruitment processes, focusing on candidates with high self-control and independence traits while developing support mechanisms for individuals exhibiting high-risk personality combinations. The risk stratification framework provides a practical tool for human resources departments to segment employees and deliver targeted interventions based on evidence-based risk profiles. Industry-specific strategies should reflect the distinct challenges identified across sectors. Banking organizations require comprehensive approaches addressing compensation transparency and management diversity, while information technology companies should focus on maintaining their advantages in flexible work arrangements while supporting employees facing innovation 136   RESEARCH, METHODS AND ANALYSIS IN SOCIAL SCIENCES . . . pressures. Manufacturing and building sector organizations should emphasize job security messaging and benefits communication to address their specific retention challenges. 10.3. Methodological Advancement and Future Applications The superior performance of survival analysis models compared to traditional binary classification approaches establishes a new standard for employee retention analytics. Organizations implementing these methodologies gain significant advantages in prediction accuracy and timing optimization for intervention efforts. The Random Survival Forest model achieving 64.1% discrimination ability represents a substantial improvement over conventional approaches and provides actionable insights for strategic decision-making. The validation results confirming model stability over time support widespread organizational adoption of these analytical approaches. Human resources departments can implement these models with confidence in their continued predictive performance and adapt intervention strategies based on evolving risk assessments. 10.4. Strategic Recommendations for Organizational Implementation Organizations should establish comprehensive talent analytics capabilities that integrate personality assessment, demographic profiling, and organizational factor monitoring into unified retention management systems. The development of early warning systems based on the identified risk factors can enable proactive intervention before employees reach critical departure decision points. Investment in coaching and mentoring programs represents a high-return strategy supported by quantifiable evidence from this analysis. Organizations should prioritize these programs in high-turnover industries and customize approaches based on specific personality risk profiles identified among their workforce. The research establishes personality-based talent management as a scientifically supported approach that can deliver measurable improvements in organizational retention outcomes. By implementing evidence-based strategies that address the complex interplay of individual differences, organizational factors, and industry dynamics, organizations can achieve sustainable competitive advantages through enhanced talent retention and reduced turnover costs. PREDICTIVE EMPLOYEE RETENTION ANALYTICS: DEVELOPING . . .   137 Future organizational success in talent management will increasingly depend on sophisticated analytical approaches that recognize the temporal nature of employee decisions and the predictive power of personality trait combinations. Organizations adopting these methodologies position themselves to achieve superior retention outcomes while optimizing resource allocation for maximum strategic impact. References Abubakar, R. A., Chauhan, A., & Kura, K. M. (2014). Role of demographic variables in predicting turnover intention among registered nurses in Nigerian public hospitals. International Journal of Business and Technopreneurship, 4(3), 373-383. Agyeman, C. M., & Ponniah, V. M. (2014). Employee demographic characteristics and their effects on turnover and retention in MSMEs. International Journal of Recent Advances in Organizational Behaviour and Decision Sciences, 1(1), 12-29. Al Hilali, K. S., Al Mughairi, B. M., Kian, M. W., & Karim, A. M. (2020). Coaching and mentoring. Concepts and practices in development of competencies: A theoretical perspective. International Journal of Academic Research in Accounting, Finance and Management Sciences, 10(1), 41-54. Al Mamun, C. A., & Hasan, M. N. (2017). Factors affecting employee turnover and sound retention strategies in business organization: A conceptual view. Problems and Perspectives in Management, (15, Iss. 1), 63-71. Andrade, J. L., & Valencia, J. L. (2021, December). Modeling lapse rates using machine learning: a comparison between survival forests and cox proportional hazards techniques. In Anales del Instituto de Actuarios Españoles (No. 27, pp. 161-183). Burk, L., Zobolas, J., Bischl, B., Bender, A., Wright, M. N., & Sonabend, R. (2024). A Large-Scale Neutral Comparison Study of Survival Models on Low-Dimensional Data. arXiv preprint arXiv:2406.04098. Chamorro-Premuzic, T., Akhtar, R., Winsborough, D., & Sherman, R. A. (2017). The datafication of talent: How technology is advancing the science of human potential at work. Current Opinion in Behavioral Sciences, 18, 13-16. Church, A. H., Rotolo, C. T., Margulies, A., Del Giudice, M. J., Ginther, N. M., Levine, R., ... & Tuller, M. D. (2015). The role of personality in organization development: A multi-level framework for applying personality to individual, 138   RESEARCH, METHODS AND ANALYSIS IN SOCIAL SCIENCES . . . team, and organizational change. In Research in organizational change and development (pp. 91-166). Emerald Group Publishing Limited. Clutterbuck, D. (2008). What’s happening in coaching and mentoring? And what is the difference between them? Dev Learn Organ 22 (4): 8–10. Conrod, P. J. (2016). Personality-targeted interventions for substance use and misuse. Current addiction reports, 3(4), 426-436. Eshet, Y., & Harpaz, I. (2021). Outstanding Employees Performance: Personality Traits, Innovation and Knowledge Management. Focbit, B. (2025). Factors influencing employee resignations: Basis for Strategic Measures to Address Turnover and Improve Retention. Business Fora Business and Allied Industries International Journal, 4(1), 39-50 Garvey, R., Stokes, P., & Megginson, D. (2008). Coaching and mentoring: Theory and practice. Germer, S., Rudolph, C., Labohm, L., Katalinic, A., Rath, N., Rausch, K., ... & AI-CARE Working Group. (2024). Survival analysis for lung cancer patients: A comparison of Cox regression and machine learning models. International Journal of Medical Informatics, 191, 105607. Hieker, C., & Rushby, M. (2020). Key Success Factors in Implementing Sustainable Mentor Programmes in Large Organisations. International Journal of Evidence Based Coaching & Mentoring, 18(2). Jha, S. (2009). Determinants of employee turnover intentions: A review. Management Today, 9(2). Kalaiselvan, K., & Nachimuthu, K. P. (2016). Strategic approach to talent management: A theoretical model. Journal of Contemporary Research in Management (JCRM), 11(1). Kerr, S. P., Kerr, W. R., & Dalton, M. (2019). Risk attitudes and personality traits of entrepreneurs and venture team members. Proceedings of the National Academy of Sciences, 116(36), 17712-17716. Madariaga, R., Oller, R., & Martori, J. C. (2018). Discrete choice and survival models in employee turnover analysis. Employee Relations, 40(2), 381395. Nagadevara, V., Srinivasan, V., & Valk, R. (2008). Establishing a link between employee turnover and withdrawal behaviours: application of data mining techniques. Neupane, R. (2015). Effects of coaching and mentoring on employee performance in the UK hotel industry. International Journal of Social Sciences and Management, 2(2), 123-138. PREDICTIVE EMPLOYEE RETENTION ANALYTICS: DEVELOPING . . .   139 Passmore, J. (2007). Coaching and mentoring: The role of experience and sector knowledge. International Journal of Evidence Based Coaching and Mentoring, 5(10). Samuel, M. O., & Chipunza, C. (2009). Employee retention and turnover: Using motivational variables as a panacea. African journal of business management, 3(9), 410. Sari, N. W. W. S., Kurniawan, A., & Ana, E. (2024). Modelıng Employee Resıgnatıon Usıng A Semıparametrıc Approach Cox Proportıonal Hazard. Barekeng: Jurnal Ilmu Matematika Dan Terapan, 18(4), 2471-2478. Silva, M. R. A., de Amorim Carvalho, J. C., & Dias, A. L. (2019). Determinants of employee retention: A study of reality in Brazil. In Strategy and superior performance of micro and small businesses in volatile economies. IGI Global Scientific Publishing. 44-56. Silzer, R., & Dowell, B. E. (Eds.). (2009). Strategy-driven talent management: A leadership imperative. John wiley & sons. Sridar, S. P. (2023). Data-driven talent strategy: Navigating workforce optimization with AI and analytics. ShodhKosh: Journal of Visual and Performing Arts, 4(1). Tarique, I. (2021). Talent management: An introduction. In The Routledge companion to talent management. Routledge. 1-6. Van den Berg, P. T., & Feij, J. A. (2003). Complex relationships among personality traits, job characteristics, and work behaviors. International Journal of Selection and assessment, 11(4), 326-339. Vlasveld, M. C., van der Feltz-Cornelis, C. M., Anema, J. R., van Mechelen, W., Beekman, A. T., van Marwijk, H. W., & Penninx, B. W. (2013). The associations between personality characteristics and absenteeism: a cross-sectional study in workers with and without depressive and anxiety disorders. Journal of occupational rehabilitation, 23(3), 309-317. Wilcox, M. (2016). Effective talent management: Aligning strategy, people and performance. Routledge. Woods, S. A., Mustafa, M. J., Anderson, N., & Sayer, B. (2018). Innovative work behavior and personality traits: Examining the moderating effects of organizational tenure. Journal of Managerial Psychology, 33(1), 29-42. Wöcke, A., & Heymann, M. (2012). Impact of demographic variables on voluntary labour turnover in South Africa. The International Journal of Human Resource Management, 23(16), 3479-3494. 140   RESEARCH, METHODS AND ANALYSIS IN SOCIAL SCIENCES . . . Wu, R., Zhao, X., Li, Z., & Xie, Y. (2024). The role of employee personality in employee satisfaction and turnover: insights from online employee reviews. Personnel Review, 53(7), 1581-1611. Zainal Abiddin, N. (2006). Mentoring and coaching: the roles and practices. Available at SSRN 962231. Zimmerman, R. D. (2008). Understandıng The Impact Of Personalıty Traıts On Indıvıduals’turnover Decısıons: A Meta‐Analytıc Path Model. Personnel Psychology, 61(2), 309-348. 141 CHAPTER IX DEVELOPMENT OF PUBLIC BANKS IN THE TURKISH BANKING SYSTEM BETWEEN 2005 AND 2024 MİNİRE KIRBAŞLI1 1(Asst.Prof. Dr.) Istanbul University-Cerrahpaşa, Vocational School of Social Sciences e-mail: [email protected] ORCİD : 0000-0001-9544-4023 1. Introduction Banks within the financial system play a key role in the country’s economic development. Particularly, public banks hold a significant share of the Turkish banking system. Three of the public banks operating in Turkey are members of the Banks Association of Turkey (TBB), while the other three are members of the Participation Banks Association of Turkey. Only public deposit banks that are members of the TBB are included in the scope of this study. In this study, the contribution of the public banks operating in Turkey (Ziraat Bank, Vakıfbank, Halkbank) to the Turkish economy between 2005 and 2024 was examined through selected balance sheet items. Participation public banks were excluded from the study due to their distinct operating structures and their inclusion in a separate category in the TBB data set. This choice was made to more clearly assess the macroeconomic impact of public deposit banks and maintain data consistency. Public deposit banks operating as members of the Banks Association of Turkey; · Ziraat Bankası A.Ş. of the Republic of Turkey was founded in Pirot in Serbia during Ottoman Empire by Mithat Pasha in 1863 and continued its 142   RESEARCH, METHODS AND ANALYSIS IN SOCIAL SCIENCES . . . activities as a local loan fund until 1888. Starting from 1888 country-wide bank by opening branches in İstanbul and other main cities. · Türkiye Halk Bankası A.Ş. (Halkbank) - It was established with the Halk Bankası and Halk Vakfı (People’s Bank and People’s Funds Law) numbered 2284, dated 1933, and started its operations in 1938. · Türkiye Vakıflar Bankası A.Ş. (VakıfBank) - was established on January 11, 1954, with a special law numbered 6219. These three banks were established with public ( state ) capital and operate in the field of retail and commercial banking services. A common characteristic of public banks is their use as a strategic tool in the implementation of public policies and development of the country. This study examines the role of public deposit banks, members of the Banks Association of Turkey, within the Turkish Banking System (TBS) and their contribution to the national economy. The analysis was based on selected balance sheet items published by the Banks Association of Turkey (TBB), and the data was evaluated in both Turkish Lira and US Dollars for five-year periods (2005, 2010, 2015, 2020, and 2024) as of December 31st. Examining data over five-year periods provides an opportunity to evaluate how the sector responds to macroeconomic conditions, regulatory changes and global financial events from a more comprehensive and strategic perspective. This type of assessment can provide a solid foundation for making more accurate predictions about the sector’s future performance. Furthermore, periodic analyses make economic cycles (growth, recession, recovery), financial crises, regulatory and policy changes, and structural transformations (technology, competitive dynamics, customer attitudes, etc.) more visible in the medium term. In the first part of the study, data analysis was conducted on selected balance sheet items of the Turkish banking system, including total assets, total loans, guarantees and under-takings, total deposits, total equity, paid-in capital and net profit/loss. In the second section, the same balance sheet items are used to examine public deposit banks. In the third section, the place of public banks in the Turkish banking system is evaluated, using selected balance sheet items and financial ratios. 2. Banking Sector In this study, the periodic distribution (TL-Dollar) of selected sector items over five-year periods as of December 31st is presented in the tables below. The PREDICTIVE EMPLOYEE RETENTION ANALYTICS: DEVELOPING . . .   143 dollar value of the data was calculated using the Central Bank of the Republic of Turkey’s foreign exchange sales rates as of December 31st. Due to the high inflationary period experienced in the Turkish economy, particularly in recent years, TL-based values reflect the effects of inflation. Therefore, dollar-based data was used for the analyzed items. Table 1: Selected Balance Sheet Items of the Banking Sector (Billion TL) Years Total Assets Total Loans Guarantees and Under-Takings Total Deposits Total Equity Paid-in Capital Net Profit/Loss for the Period 2005 397 153 61 254 54 20 6 2010 962 509 142 615 129 45 21 2015 2,236 1,459 423 1,251 252 65 26 2020 5,664 3,609 938 3,308 572 111 55 2024 29,900 15,874 5,625 18,305 2,684 327 596 Source: TBB, https://www.tbb.org.tr/istatistiki-raporlar, Access Date: Edited using data from 15.05.2025. Table 2: Central Bank of the Republic of Turkey Exchange Rates Year Foreign Exchange Sales Rates 2005 1.3483 2010 1.5450 2015 2.9233 2020 7.4327 2024 35.3438 Source: TCMB, https://www.tcmb.gov.tr/kurlar/kurlar_ tr.html, Erişim Tarihi: 15.05.2025 Note: The above data is based on the Indicative Quality Data Set at 15:30 on December 31st. 144   RESEARCH, METHODS AND ANALYSIS IN SOCIAL SCIENCES . . . Table 3: Selected Balance Sheet Items of the Banking Sector (Million $) Years Total Assets Total Loans Guarantees and Under-Takings Total Deposits Total Equity Paid-in Capital Net Profit/Loss for the Period 2005 294,423 113,520 45,147 188,073 39,855 14,864 4,238 2010 622,573 329,361 91,955 397,852 83,551 29,028 13,825 2015 764,887 498,928 144,684 427,838 86,072 22,093 8,772 2020 761,976 485,524 126,166 445,088 76,955 14,873 7,371 2024 847,769 449,123 159,149 517,912 75,949 9,256 16,870 Source: Edited by using Table 1 and Table 2. 2.1 Total Assets It is observed that the total assets of the banking sector increased by approximately 75 times (7,748%) from 397 billion TL in 2005 to 29.9 trillion TL by the end of 2024. When the total assets of the sector is examined in dollar terms, it is observed that it has grown approximately 2.9 times from 294 billion dollars to 848 billion dollars in 2005. While the increase in total assets from $294 billion to $765 billion between 2005 and 2015 indicates a strong increase, the fact that it decreased from $765 billion in 2015 to $762 billion in 2020 suggests that there may be a stagnation in real growth. However, its rise to 848 billion dollars after 2020 can be considered a recovery period. 0 100,000 200,000 300,000 400,000 500,000 600,000 700,000 800,000 900,000 2005 2010 2015 2020 2024 Million $ Graph 1: 5-Year Periodic Change in Total Assets in the Sector (Million $)