Asymmetric Vacillation in the FMCG Industry: A Case Comparison of Procter & Gamble and Unilever

: While existing studies regarding organizational vacillation theory have focused on examining how vacillation may lead to ambidexterity, few studies have focused on how vacillation itself happens and whether it happens symmetrically or not. To bridge this research gap, this paper analyzed organizational vacillation over time with two canonical cases while examining patterns of organizational structure over time. Unlike previous studies that only revealed the existence of vacil-lation between centralization and decentralization, this study revealed that vacillation is observed with an asymmetric ratio of duration in the business world by ﬁnding that each company within the same industry spends a greater portion of time in a certain organizational structure than the other. By analyzing these changes throughout the business history, this study found that organizational vacilla-tion happens asymmetrically while alternatively shifting between centralization and decentralization. Based on the case study result, this study draws propositions that can enable future researchers to advance theoretical and empirical understanding toward asymmetric vacillation.


Introduction
The fast-moving consumer goods (FMCG) industry includes a wide variety of packaged goods for consumers in food and beverages, cosmetics, household goods, and baby and feminine care products. Unlike in the past when firms only had to survive within limited geographic boundaries, globalization has significantly impacted global economic strategy (Jones 2002). Thus, globalization has resulted in boundless competition among existing firms. Now, firms must strive to adopt different strategies to maintain their competitive advantage by differentiating themselves from their competitors (Oraman et al. 2011). Competition within the FMCG industry is accelerating because it is relatively easy for new entrants to enter the market since it is not technology-and capital-intensive compared to other industries. As a result, the FMCG industry is already saturated with not only multinational conglomerates such as Procter & Gamble, Unilever, and Johnson & Johnson but also with local players in each market. Due to globalization and intense competition, Procter & Gamble and Unilever have both constantly shifted their organizational structure between centralization and decentralization to optimize their operations while discovering new market opportunities around the world. This study contributes to organizational studies by examining how two FMCG giants, Procter & Gamble and Unilever, have been at the forefront of the market by asymmetrically vacillating their organizational structures throughout the decades.
The central goal of our research is to examine the dynamics of organizational vacillation between centralization and decentralization over time and to investigate why one firm maintains a certain organizational structure whereas another firm maintains a different organizational structure under similar situations. To reach our research goal, we first reviewed the existing literature regarding organizational vacillation and went over the open innovation literature to enable a deeper understanding of organizational structures. Then, we conduct a comparative study on the case of Procter & Gamble and Unilever, as these firms ranked first and second in selling varied items including personal care, cosmetics, baby & feminine care, and cookware products.
By analyzing both companies' histories regarding the changes in organizational structure, this paper reveals that even companies in the same industry show asymmetric vacillation over time. Based on our analysis, we aim to theoretically explain and illustrate how our observation contributes to organizational vacillation theory. To be specific, by comparing two familiar cases, we attempt to propose that organizations vacillate between two different structures asymmetrically. Unlike the existing studies that revealed merely the existence of organizational vacillation and revealed that organizations vacillate alternatively between centralization and decentralization, this study adopted the concept of 'time duration' to examine whether organizations spend an equal amount of time in both structures. By revealing an asymmetric pattern of organizational vacillation, this study can expand organizational vacillation theory in the future by offering another important boundary condition between organizational vacillation and firm performance. This can offer important insights to practitioners since leaders who are willing to enhance their organizational performance by skillfully vacillating between different structures should understand that vacillation tends to occur asymmetrically and what motivation and effects this asymmetric vacillation can have.
We clearly acknowledge that the results of a case study using two examples cannot be generalized since empirical testing is missing. However, the aim of this paper is not to empirically test the asymmetric vacillation but rather to depict a novel phenomenon that current theory does not focus on. Thus, we hope that by focusing on the vacillation phenomenon itself and revealing that organizations vacillate unevenly between different structures, this paper can contribute to existing vacillation theory.

Literature Review and Research Question Organizational Vacillation Theory
Despite the long-standing argument that organizational changes are a simple response to external changes, Nickerson and Zenger (2002) and Brown and Eisenhardt (1997) argued that the organizational choice regarding form and structure is not fully understood as a discrete response toward the environment, rather focusing on continuous changes and the modulation between the choices, which is known as vacillation. A metaphrase of vacillation refers to the state of irresolution or unstable movement and fluctuation (Nickerson and Zenger 2002).
Prior studies describe the phenomenon of organizational vacillation as the "timehonored cycle between centralization and decentralization" (Dougherty et al. 1993, p. 127). It has also been described as "Companies oscillating between centralization and decentralization." (Cummings 1995, p. 112). Scholars such as Mintzberg expressed the vacillation between centralization and decentralization as the "swings" (Mintzberg 1979). This attention toward the vacillation phenomenon continues to this day through studies on more detailed frameworks of vacillation such as interval, scale, and its influence on performance.
According to Nickerson and Zenger (2002), firms make decisions to oscillate between or among each different structure to certain, yet temporary, levels of functionality that are unachievable when organizations remain with one particular structure under certain conditions. Therefore, firms generate value from the ambidextrous organization that temporarily emerges from vacillation. Their research refined the contingent-fit-based theories of organizational alignment, especially for the static world. As a result, the study revealed that even a static environment demands a more dynamic response from an organization. Moreover, their model suggests that when no there is no exact organizational setting that indicates a perfect match to exchange attributes, managers can still generate profits by adapting organizational models to improve the functionality of a firm to the optimal level, while keeping the cost of the change sufficiently low. They also revealed that exploration and exploitation require opposite organizational designs for a firm's optimal functioning (Nickerson and Zenger 2002). For instance, centralization promotes exploitation while decentralization promotes exploration; thus, firms strive to dynamically achieve high levels of both structures and sequentially alternate these structures.
Drawing upon the vacillation theory, Kang et al. (2017) measured exploration and exploitation and provided empirical evidence regarding the impact of organizational vacillation on performance. The result of this study revealed that there were inverted U-shape relationships between vacillation frequency, scaled vacillation, and firm performance while using patent-related measures of exploration and exploitation. The study of Nickerson and Zenger (2002) contributed a theory of dynamic organizational choice and identified how organizational structure can be changed even in a static environment, and Kang et al. (2017) supplemented Nickerson and Zenger's study by providing empirical evidence, but still existing accounts fail to answer when such conditions should be held.
Based on organizational vacillation theory, recent studies have examined more diverse effects of vacillation on firm performance. Kang et al.'s (2017) study examined two different types of vacillation between exploitation and exploration based on the radicalness of the transition in firm performance. According to the study, an incremental transition had a positive effect on firm performance; however, a discontinuous, quantum transition had a negative effect on firm performance. Luger et al.'s (2018) study approached vacillation from a contextual perspective by suggesting boundary conditions where reinforcing ambidexterity can have both advantages and disadvantages toward organizational performance. Not have only empirical tests but also a systematic review been conducted (Kassotaki 2022;Rojas-Córdova et al. 2022). Rojas-Córdova et al. (2022) reviewed 20 years of studies that examined exploration and exploitation and how organizational vacillation theory has evolved during the past decades. Researchers such as Zimmermann et al. (2018) approached vacillation and ambidexterity from a managerial perspective and examined whether frontline managers will assign great importance to senior managers' initial structural design choices by adopting an inductive study.
The existing literature has found that organizations change their structures alternatively between centralization and decentralization; however, the literature has yet to find an answer as to whether organizational vacillation happens symmetrically or not, and if so, under what circumstances an organization remains in a specific structure for a longer time. In other words, based on the existing studies that found that organizational vacillation happens alternatively, this study aims to find out whether this vacillation will happen symmetrically. We attempt to fill this gap in organizational research by offering a study that answers the question of whether organizations stay in a specific structure for a longer time and why this may occur. To achieve this goal, this study examines the changes in organizational structure between centralization and decentralization over time in Procter & Gamble and Unilever. Comparing these two FMCG behemoths can offer great chances to compare two familiar examples, which can permit a more robust discussion regarding 'asymmetric vacillation' in the future.

Methodology
A fundamental aim of this paper was to examine the dynamics of organizational vacillation between centralization and decentralization over time to find whether organizational vacillation happens symmetrically or not. To answer our research question, an in-depth case analysis method was chosen since the concept of organizational vacillation and organizations' changes toward centralization and decentralization can be better shown by illustrating and comparing two cases in the literature rather than by empirically measuring these changes. Boumgarden et al. (2012) argued that vacillation is episodic, which in turn requires detailed case studies to identify such episodes. Similarly, Klein's (2000) study adopted a GM-Fisher Body case to provide a vivid demonstration of how the operation of the long-term contract was initiated, and Boumgarden et al. (2012) and Nickerson and Zenger (2002) also conducted a case study of Hewlett-Packard in order to illustrate organizational vacillation. Accordingly, we also adopted a case study method to approach our research question.
In this article, we first developed a more concrete history of Procter & Gamble and Unilever by drawing on a wide array of data sources including annual reports, letters to investors, and press releases. A comparison between two multinational companies provides a better-understood set of facts and enables future discussions about the antecedents or determinants of asymmetric swings between centralization and decentralization within an organization. In this section, we first analyze how each company's organizational structure has changed up until 2020 based on a wide array of data sources. We examined both companies over a 20-year period of time, during which Procter & Gamble and Unilever competed for both first and second place in the consumer goods industry around the world. The specific reason that we conducted comparative research for the last two decades is that most of the studies on FMCG companies are outdated and the data for comparing both companies are only available from 2000. Our analysis included various news articles, business case studies, and media conferences of CEOs from both companies, and it heavily relied on annual reports and letters to investors.
Using these data sources, we constructed a focused history of Procter & Gamble and Unilever based on the significant shifts in their organizational structures. We approached our research questions by examining the structural shifts of each organization over time.
Our study method was based on the theory-building approach by Anderson (1983) and Pinfield (1986), who used case studies as the primary approach.
We examined a time during which both companies became the largest and the most profitable FMCG companies globally. Based on various data sources, we focused on the history of the two companies between 2000 and 2020, and we organized this into two sections: centralization and decentralization. Significant shifts within the organizational structure were examined during the time we selected.
Then, we categorized each firm's organizational structure into either centralization or decentralization based on the following indicators that appeared on press releases and annual report statements. These included statements or expressions such as "standardized overall processes", "simplification", "standardization", and "focused on centralization" and were regarded as clear evidence of centralization. On the other hand, the following statements or expressions such as "gave regional officers more autonomy", "focus more on emerging markets", and "decentralization" were regarded as evidence of decentralization. Sometimes different evidence was presented for each organization's business units (i.e., centralization for marketing and decentralization for R&D, sales, etc.), but we prioritized compiling multiple pieces of evidence to identify the "overall" and "dominant" strategic structure of the organization as a whole. Specific keywords or phrases used as evidence for determining whether each organizational structure was centralized or decentralized in the data analysis work are summarized in the following table (Table 1). • "Part of that effort is centralizing warehousing and distribution facilities" (Tweh and Coolidge 2014) • "Consolidating its HQ will allow Unilever to sharpen its central functions" (Friesl 2018) • "P&G boss targets leaner and more decentralized business, in first public address" (Pitman 2016) • "For much of its past, Unilever was a 'federation' of highly localised businesses that benefited from high levels of autonomy" (Friesl 2018) Major keywords In 1837, a candlemaker named Procter and a soap maker named Gamble founded the P&G (Procter & Gamble) Company in Cincinnati, Ohio. Since Cincinnati was one of the largest meatpacking centers in the country, P&G was able to access high-quality animal fat at a reasonable price, which was a primary raw ingredient for both candles and soap. During the civil war, P&G supplied its products to the Union armies, and its product's high quality and affordable price made the organization nationwide wellknown company when the soldiers finally returned home after the war . From the beginning, P&G gave importance to product innovation. Norris James Gamble, Gamble's son, was also a proficient chemist, and he developed the first soap made in the United States, which was equivalent to European high-quality imports. James continued to transform P&G's production processes from an art to a science with help from his professors (Piskorski and Spadini 2007). These efforts toward scientific production processes finally led to the large-scale production of Ivory in 1887. Finally, in 1890, P&G became an industry pioneer in centralizing R&D labs (Procter & Gamble 1999). Based on the centralized R&D labs, P&G continuously diversified its product lines within fast-moving consumer industries, such as laundry detergents, paper products, kitchenware, and personal-care products. As the product lines expanded and the business portfolio became more complex, it was essential for P&G to partition each brand manager to target different consumer segments based on which product lines they were in charge of. In the process, the organization started forming around the product lines to make more consumerfocused business decisions and thus created its first product-category division. Although the company seemed decentralized with lower-class managers having the authority to make decisions, the company's core functions, such as R&D and manufacturing, remained centralized (Piskorski and Spadini 2007).
Starting in 1948, P&G constantly built its foreign businesses based on its strong presence in the United States. For the United States, where customers share great homogeneity with enormous market opportunities, P&G developed product division management to effectively manage various product lines. At this time, the core two dimensions of the organization were functions and brands (Piskorski and Spadini 2007). With this initial structure, brand managers who worked for the same product division competed in the marketplace while sharing resources with other divisions. After, P&G made changes from its product division-centered structure to enabling each unit to obtain its own functions including manufacturing and sales. This structural change made almost 40 different business units.
Unlike in the United States, which was a homogeneous market, P&G had to tailor its products and operations to fit the local needs and tastes for Western European countries. Even though local managers in each country evaluated and transformed their products accordingly, the core technologies of the products were still based on the centralized R&D labs in the United States. Thus, the company's core technologies were still concentrated in the United States lab, and just incremental adaptations were then made to suit the taste of local customers and law requirements. During this time, country managers rather than brand managers were in charge of profitability and market strategy. This existing structure led to a complete separation between the two markets, and thus huge impasse was presented between European segments and headquarters in the United States. To make matters worse, structure at this time lacked standardization which made each country to come up with its own manufacturing processes. All these impediments made the organization's overall operations unstable and unprofitable, and finally P&G realized the importance of centralization and thus started to redesign overall organization's management system. From 2000 to 2005, the fundamental question for Procter & Gamble was whether the matrix of its structure was well-aligned in the long term. Because the previous scenario could not fully handle the conflict between different cultures and processes, many experts identified it as a problem. Furthermore, as P&G continuously diversified, an unparalleled and unnecessary number of general manager positions were generated based on its matrix. Thus, in September 1998, P&G finally announced its plan to initiate its "Organization 2005" program, which was anticipated to take six years to restructure the overall business structure and operations. This initiative resulted in huge changes. From a management perspective, redundant six layers of management were eliminated to ensure the efficiency of the operations. Moreover, under the desire for globalization, this program radically changed its matrix and replaced it with three new business units: Global Business, Market Development Organizations, and a Global Business Services unit (Piskorski and Spadini 2007). The Global Business Unit was primarily in charge of products, Market Development Organizations were responsible for targeting markets and initiating expansion plans, and the Global Business Services unit managed internal operations (Procter & Gamble 2001). Given the background of this program's implementation, it is natural that the main focus of the program was to reduce costs through simplification of the product lines and standardization.
The annual report from this time stressed the importance of implementing its own program which aimed to achieve simplification and standardization of the overall organizational structure. This program included projects across the whole organization from consolidating selected manufacturing infrastructures, redesigning organization, simplifying existing products, and divesting brands that were not strategically well aligned (Procter & Gamble 1999, p. 14). As announced in the annual report, P&G implemented the Organization 2005 program with a specific focus to mainly standardize and centralize its manufacturing processes and change from a geographic focus to product-based functions. This program was initiated effective from 1 July 1999. This restructuring brought technology improvements including the standardization of overall operations and manufacturing processes, which resulted in the ability to streamline operations and improve manufacturing productivity, which helped to reduce costs.
To implement the wide, extensive restructuring program of Organization 2005, on January 1, Durk Jager was appointed as the new CEO. His goal was to use the Organization 2005 program to provide P&G with a groundbreaking brand based on innovative technologies rather than incremental improvements of existing technologies (Brown and Anthony 2011). Fortunately, his efforts quickly paid off. Right after implementing the Organization 2005 initiative, P&G's first fiscal results in 1999 showed an immediate improvement in business performance. According to the annual report, the company's overall sales, core net earnings, and annual revenue showed marked improvements.
Yet, the glory did not last long. Unfortunately, the situation started to deteriorate dramatically on 7 March 2000. Revenue increased compared to 1999; however, the core earnings decreased by ten percent compared to the financial performance from January to March. P&G stated that this resulted from various factors including increased costs of raw materials and delays in FDA approvals. P&G continued to also struggle with increased competition, lower volume growth, and negative currency effects, which led P&G to lose its market share in fourteen categories. Continued poor performance made Jager have no choice but to resign (Piskorski and Spadini 2007).
After Jager's resignation, Lafley was appointed as the next CEO to revitalize the organization, leading the company for all of the 2000s. During the early stages of having a new CEO (from 2002 to 2003), P&G was still implementing its restructuring program that began in 1999. The company started to realize the importance of developing markets including China or eastern Europe, but still, the major focus was on technology improvements as well as the standardization of manufacturing and other work processes. During this stage, P&G strived to tightly control manufacturing and product costs by standardizing its manufacturing platforms. The annual report from 2004 states the organization's plan to improve manufacturing productivity while saving capital investment in order to meet new consumer demand. Achieving such cost savings enables the organization to keep kneen consumer value and to invest more in new products (Procter & Gamble 2004).
Moreover, most market sales were still generated from North America, where P&G was already a dominant player. Even though the company started to notice new market opportunities, its annual report from 2002 and 2003 still highlights the importance of expanding into these new markets with standardized products and technologies since consumers in these markets are expected to have homogeneous performance expectations as the developed market (Procter & Gamble 2003). Expanding new markets with standardized products and technologies will be effective since consumers in these new, developing markets have far less purchasing power.

•
Epoch 2: Decentralization (2006)(2007)(2008) From 2006 to 2008, P&G's main strategic focus moved to decentralization. In this period, the company's strategic focus was to decentralize overall business functions rather than centralization, which was a noticeable change compared to the previous years. It continuously highlighted the importance of customization, developing markets, and geographic divergence. P&G's performance in the fiscal year of 2006 demonstrated the company's ability to sustain its growth. Every geographic region showed great sales vol-ume and growth, which were particularly led by emerging countries' sales growth (Procter & Gamble 2006). It also stated that there were growth opportunities of more than USD 20 billion over the next several years through the expansion of the organization's category presence in developing markets. In 2007, P&G implemented the Go-to-Market Reinvention program (G2MR), which focused on making incremental improvements in supply chain efficiency with an increased focus on prioritizing retail customers in emerging markets such as India and Brazil.

•
Epoch 3: Centralization (2009-2015) Since 2009, P&G has tried to rejuvenate its productivity based on the term "digitization". With digitization, P&G's focus was to standardize, automate, and fully integrate its existing systems and data so that it could be reactive to specific market situations and to enable real-time operating and decision-making. Over the next four years, it consolidated planning tasks and implemented new simulations. This tool to re-centralize and digitize improved P&G's efficiency of the manufacturing process and created a sustainable organizational structure. P&G attempted to simplify the planning processes and implemented new technologies that could lower transportation and warehousing costs and thus improve productivity by 25% or more by centralizing the overall operating systems.
The annual report from the year 2013 states that by implementing a restructuring program, it will become a simpler and more concentrated organization while only focusing on 70 to 80 brands (Procter & Gamble 2013). Under this restructuring process, unproductive and non-strategic brands, product lines, products will be divested. By pursuing centralization again, P&G eliminates all the redundant functions of its operation and to standardize, simplify the organization for faster and more effective innovation. This program was executed across virtually all of its Market Development Organizations (MDO) and Global Business Units (GBU).
These efforts continued until the year 2015. The annual report at this time stated highlights its efforts to put the strategies and resources in order to make P&G a more faster-growing organization while having a simpler organizational structure and much more focused and concentrated portfolio" (Procter & Gamble 2015).

•
Epoch 4: Decentralization (2016-2020) Starting from the year 2016, a new CEO named Taylor started his term in the organization. Previously, P&G had focused on centralizing its operating system and reducing costs, but in 2016, P&G started to decentralize its businesses. Taylor strongly emphasized the importance of decentralization in his inaugural speech at a New York conference by admitting that the company was too centralized which made it hard to address new, emerging market opportunities, and that from now, the organization will empower regional managers in order to have direct relationship with the market and adapt faster into the new market (Swamynathan 2016).
Continuing, Tamer Younes, an executive board member and the Head of Corporate Affairs at Procter & Gamble, said that his company would focus more on reforming the organization in order to expand and localize its production rather than standardizing its value chain, with the aim to rank P&G as one of the top five exporters in the industrial chemicals sector (Omran 2018). As a result, between 1999 and 2020, P&G's organizational structure was centralized for thirteen years, while it was only decentralized for seven years.

The Business History of Unilever before 2000
Unilever was founded in 1929 as a combination of two existing companies, Unilever Ltd. (which changed to plc later) and Unilever NV, with each company having the same amount of claim on the profits. Unilever's legal and financial operations spanned over two Western countries, the Netherlands and the United Kingdom. In this phase, Unilever also underwent the acquisition of several small operating companies that had been held by separate Dutch and English companies. Through this merger, Unilever became the largest firm in Europe. While becoming the largest firm in Europe, it had the intention to operate as one integrated business, not as a combination of several separate companies. Therefore, for Unilever, this was a vital moment that would decide which organizational structure would be supportive for upcoming business situations and strategies.
The result was a decentralized structure at the continent level. After 1929, as the political and economic flow of European integration had reached its optimal level, Unilever's support for European integration began to show in various European markets. Unilever itself attempted to expand its geographical presence beyond its primary market, which was centered mostly in the northwest part of Europe. Thus, Unilever aimed to expand to Italy and later to many other southern parts of Europe. To integrate the business process of each nation and achieve considerable economies of scale, manufacturing and distribution were set to be arranged regionally. In 1958, a report written by one of the senior marketing managers stated that the European integration initiative widened the possibility of a more international approach to produce and distribute its packaged goods throughout the whole continent.
Its decentralized structure strengthened after the outbreak of the cold war and economic nationalism during the 1930s. Therefore, Unilever's post-war organizational structure was closer to the decentralized structure at the national level. This was mainly because the firm had to depend on local initiatives to get back on its feet, as businesses had been devastated in many European countries. Local-level decision-making and business processes helped the business selling consumer goods to adapt to the tastes, habits, and legislation of local markets, which were divergent. Moreover, these characteristics of a decentralized structure were created naturally and grew through mergers and acquisitions. Managers of each country dealt with the responsibility for profits and operational management, which was then reported to a Rotterdam-based management group, the Continental European Group.

Organizational Vacillation of Unilever after 2000
• Epoch 1: Regionally decentralized structure begins (2000)(2001)(2002)(2003)(2004) By the early 2000s, Unilever was determined to regain its momentum by greatly changing the state of the firm. As the internet rapidly gained its popularity and stocks related to telecoms were encouraging investors to stay away from old economy stocks, Unilever began to struggle. Unilever realized the vital need to dismantle its existing organizational structure and overall operations to directly face the challenges and opportunities brought about due to the changing market conditions. Finally, in February 2000, Unilever announced a EUR 5 billion five-year initiative regarding growth strategy, which was aimed at bringing about a significant improvement in its performance.
While divestment of the brand and simplification of the business process took place on the product category side to leverage economies of scale, Unilever's performance was organized into both a region and a product category. Effective from January 2001, two globally integrated divisions were Set-Foods, and Home and Personal Care (HPC). The global divisions' operations were divided into regional-based organizations, except for certain products. This decision was aimed at meeting the everyday needs of people everywhere. Unilever began forming its business reputation to one that valued different local cultures and markets all around the world. These global divisions' operations were organized into businesses on a regional basis, enabling quick responses to the needs of local markets while leveraging economies of scale from the consumer groups of one integrated region.

•
Epoch 2: A centralized Unilever: The One Unilever Program (2005Program ( -2009 By 2004, the Path to Growth program officially ended. In 2005, there was significant organizational change involving the delayering of the organizational reporting system between the boards and the operating units and clearly distinguishing the responsibility for brand development and category management into one segment and country and customer development into another. On 10 February 2005, Unilever decided that it would change the direction of its organizational structure from a decentralized structure to a simpler structure. According to its official statement, the existing regionally divided business groups were managed with each president for each integrated centralized business region unit. Moreover, a centralized system for new categories to develop category-intensive strategies and brand positioning started in April 2005. Also, the One Unilever Program was officially launched in 2005. The purpose of this program was to simplify the business, delivering cost savings and enabling Unilever to leverage global scale through three regional teams and a more globally integrated structure in several main functions (Chandola 2016) The Perennial Issue of Adaptation or Standardization of International Marketing Communication: Organizational. The program was maintained until 2009, when it officially ended. It resulted in leveraging the scale of the centralized structure with the global procurement system. Unilever integrated regional sourcing organizations across each geographical business group and further streamlined the organizational structure. It was able to create a new global customer supply structure and newly launched the customer innovation centers and the Chief Global Supply Chain. Unilever also created a centralized system (UES) for several business value chain activities such as HR, finance, IT, and information management services.

•
Epoch 3: The move back to decentralization (2010-2020) As described in Epoch 2, Unilever achieved its expected advantages in supply chain functions through a centralized structure, leading to a pursuit of a centralized structure specifically for supply chain functions. However, to meet the local demands of global consumers, Unilever set up what was called a "segmented supply chain". The segmented supply chain had specific channels, such as drug stores, each characterized by a different approach. In the area of beauty products, for example, Unilever created different supply chains to improve innovation. Therefore, Unilever varied its organizational configuration to respond faster and to be able to utilize different asset bases in the segmented supply chains.
During Epoch 2 (2005-2009), Unilever achieved financial improvement, as its profit margin and turnover recovered from the continuous decrease from 2000 to 2004. However, for the core value chain processes, like the manufacturing process, a centralized structure was not appropriate to handle the changing environment. Since Unilever produced the majority of its products through a network of more than 240 manufacturing sites around the world, Unilever invested significant amounts of money in manufacturing sites to achieve a decentralized network, enabling Unilever to maximize the global scale of its operations. Despite this slight modulation toward centralization in the manufacturing process, the priority was still on pursuing the agility to meet local demands. Therefore, this made Unilever vacillate back into a decentralized structure once again.
One of the examples showing these tendencies can be found in the Connected 4 Growth Program in 2016; the manufacturing process was conducted in a digitalized and data-driven way, pursuing a decentralized structure that was more responsive to fast-changing consumer trends. Chief financial officer Graeme Pitkethly said that Unilever was "launching more local innovations than ever before" (Gwynn 2017). Unilever allowed for more independence to local teams in order to deliver local innovations. In addition, the new Country Category Business Teams (CCBTs) were implemented and helped to make the innovation pipeline stronger. The CCBTs allowed Unilever to launch global innovations faster and provided agility in responding to local levels of micro-trends. Each division was able to allocate its assets and resources more dynamically between locations and make investment decisions based on strategic objectives. Also, in terms of organizational change, Unilever was able to deploy more resources for global brand communities and local operations. This allowed Unilever to empower employees to take more effective action locally.

Comparing the Two Cases
Over 20 years, both organizations pursued a remarkable pattern of restructuring, and whenever they pursued structural shifts from one to the other, the structural change seem-ingly focused on correcting or overcoming the challenges of the past structure. For example, P&G pursued centralization such as product standardization mostly in order to enhance organizational efficiency and improve productivity. When this tendency toward centralization became too excessive, the managerial aim shifted from centralization to decentralization and aimed to operate divisions autonomously and provide regional managers with more direct ownership so that the organizations could address market opportunities more evenly. Again, when these autonomous operation systems yielded excessive redundancies, which in turn increased the bureaucratic costs, P&G shifted from decentralization to centralization to reduce its own inefficiencies. Similarly, Unilever adopted decentralization when it aimed to prioritize meeting local needs while customizing its products into local cultures and markets while pursuing centralization when it reorganized its overall structure to simplify businesses and deliver cost savings with a more integrated structure. This finding coincides with the existing studies that illustrate that organizations tend to vacillate between centralization and decentralization alternatively.
Based on the existing studies, however, this paper sheds light on organizational vacillation theory by showing that vacillation occurs asymmetrically. Regarding our research question, the comparison between the two cases clearly showed that the organizations vacillated alternatively between centralization and decentralization; however, they tended to adhere to one specific organizational structure longer than the other. Unlike the existing studies that only revealed the existence of organizational vacillation and heavily focused on how this vacillation affects organizational performance or how this vacillation can lead to ambidexterity, this study focused on organizational vacillation itself. Based on our analysis, P&G stayed in a centralized structure for thirteen years while Unilever spent only five years. On the other hand, P&G stayed in a decentralized structure for seven years while Unilever spent fifteen years. Regarding our research question, even though both organizations made structural shifts alternatively, P&G adhered to centralization for a longer time while Unilever adhered to decentralization for a longer time. Figure 1 not only shows the organizational structure trends but also implies the concept of duration in vacillation. Even though the two companies compete in the same industry, have similar product categories, and operate their business on a similar international scale, they showed asymmetric vacillation durations. Figure 1  Also, over a 20-year period, each company made several big structural shifts between centralizing and decentralizing major decision processes and business activities. While recognizing the clear limitations of such an analysis, it is interesting to examine the pattern of the two companies' stock prices in conjunction with these structural shifts. Figure 2 graphically displays the pattern of the ratio of P&G and Unilever's daily closing stock price. As the theory of endogenous organizational change was suggested to have empirical traction in Nickerson and Zenger's research, what is evident in this chart is a cyclical pattern of P&G and Unilever's share price performance, which coincides with these changes. Also, over a 20-year period, each company made several big structural shifts between centralizing and decentralizing major decision processes and business activities. While recognizing the clear limitations of such an analysis, it is interesting to examine the pattern of the two companies' stock prices in conjunction with these structural shifts. Figure 2 graphically displays the pattern of the ratio of P&G and Unilever's daily closing stock price. As the theory of endogenous organizational change was suggested to have empirical traction in Nickerson and Zenger's research, what is evident in this chart is a cyclical pattern of P&G and Unilever's share price performance, which coincides with these changes.

Discussion
Based on a longitudinal comparative case analysis, this study analyzed how the two FMCG giants of P&G and Unilever's organizational structures vacillated over time. This study indicated that the two firms vacillated their organizational structure "asymmetrically" between centralization and decentralization. Our paper proposed the notion of 'asymmetric vacillation', which presumes that even though organizations alternatively switch between centralization and decentralization, as previous studies have revealed, each organization has its own dominant structure such that one organization stays in this dominant structure for a longer period of time than the other. It is important to note that each organization has its own dominant structure, since this can reveal the diverse meaning of ambidexterity for each organization. Asymmetric vacillation might be another reason why most organizations fail to strike a balance between centralization and decentralization or exploration and exploitation. Moreover, because organizations have their own dominant structure, it might be reasonable for some organizations to stay in one specific structure while exploiting the other structure periodically rather than exploiting both structures to the exact same extent, since switching the structures inevitably entail costs. Considering this asymmetric vacillation can expand the current vacillation theory by examining the different meanings of ambidexterity to each firm based on each firm's dominant structure. Moreover, if future studies can examine what causes this asymmetric vacillation, it will also be possible to shape organizations that possess ambidexterity from their creation.
Building on the reviewed literature and our content analysis of P&G and Unilever cases, several lessons could be learned. In order to conceptualize 'asymmetric vacillation' and to enable future studies to empirically test this interesting phenomenon, the following propositions are presented as they were drawn from the theoretical lessons of this study. We hope that the presented inductive propositions based on the case study method can enable future studies to develop our propositions and to empirically test asymmetric vacillation.

Proposition 1. While an organization vacillates between centralization and decentralization over time, an organization will adopt one of them as a dominant structure.
Organizations tend to vacillate between two opposing structures, i.e., centralization and decentralization. While the existing studies focused on the effect of having a balance between these two different structures, which is called 'ambidexterity' (  When comparing the stock price fluctuations of P&G and Unilever summarized in Figure 2 with Figure 1, it can be inferred that the stock price mostly followed the organizations' structural changes. To put it differently, whenever each organization implemented a structural shift between centralization and decentralization, there was a spike. For instance, based on Figure 2, P&G's stock price bounced back between 2008 and 2009, and this is the time when it shifted its organizational structure from decentralization to centralization based on Figure 1. During this time, P&G implemented "digitization" to integrate and centralize its business in order to enhance its overall business productivity.
Similarly, Unilever's stock price fluctuations showed a huge bounce back between 2009 and 2010 based on Figure 2, and this is the time when it shifted its structure from decentralization to centralization based on Figure 1. After shifting to a centralized organizational structure, the organization enhanced its overall cost savings. Even though this pattern by no means eliminates alternative explanations and does not guarantee the causal relationship based on an empirical examination, our observation shows that these changes were motivated by logical organizational consequences of the prior structural choice. By linking Figures 1 and 2, we found that the major structural shifts between centralization and decentralization coincided with each company's stock price fluctuations.

Discussion
Based on a longitudinal comparative case analysis, this study analyzed how the two FMCG giants of P&G and Unilever's organizational structures vacillated over time. This study indicated that the two firms vacillated their organizational structure "asymmetrically" between centralization and decentralization. Our paper proposed the notion of 'asymmetric vacillation', which presumes that even though organizations alternatively switch between centralization and decentralization, as previous studies have revealed, each organization has its own dominant structure such that one organization stays in this dominant structure for a longer period of time than the other. It is important to note that each organization has its own dominant structure, since this can reveal the diverse meaning of ambidexterity for each organization. Asymmetric vacillation might be another reason why most organizations fail to strike a balance between centralization and decentralization or exploration and exploitation. Moreover, because organizations have their own dominant structure, it might be reasonable for some organizations to stay in one specific structure while exploiting the other structure periodically rather than exploiting both structures to the exact same extent, since switching the structures inevitably entail costs. Considering this asymmetric vacillation can expand the current vacillation theory by examining the different meanings of ambidexterity to each firm based on each firm's dominant structure. Moreover, if future studies can examine what causes this asymmetric vacillation, it will also be possible to shape organizations that possess ambidexterity from their creation.
Building on the reviewed literature and our content analysis of P&G and Unilever cases, several lessons could be learned. In order to conceptualize 'asymmetric vacillation' and to enable future studies to empirically test this interesting phenomenon, the following propositions are presented as they were drawn from the theoretical lessons of this study. We hope that the presented inductive propositions based on the case study method can enable future studies to develop our propositions and to empirically test asymmetric vacillation.
Proposition 1. While an organization vacillates between centralization and decentralization over time, an organization will adopt one of them as a dominant structure.
Organizations tend to vacillate between two opposing structures, i.e., centralization and decentralization. While the existing studies focused on the effect of having a balance between these two different structures, which is called 'ambidexterity' (Andriopoulos and Lewis 2009;Junni et al. 2013;O'Reilly and Tushman 2011;Raisch et al. 2009), we suggest a different perspective: asymmetric vacillation. Both P&G and Unilever rank first and second within the FMCG industry and share great similarities regarding their product categories and operating countries; however, each organization showed differences in their dominant organizational structure. Based on this finding, we suggest that each organization will have its own dominant structure, where one tends to stay for longer period than the other while focusing on this imbalance between centralization and decentralization. We refer to this imbalance as asymmetric vacillation.
Then, we move on to understand 'why' organizations tend to have one dominant structure. Even though examining the possible antecedents that caused the asymmetric vacillation was not the main focus of this paper, the following paragraphs try to provide crude explanations of why the two organizations showed different vacillation patterns.
There are possible preceding factors for why P&G and Unilever showed different vacillation phases despite their commonalities, such as inter-industry characteristics and external situations. As was described in the introduction, they are both ranked as the first-and second-largest companies in the fast-moving consumer goods industry, which is characterized by multiple products and multiple operating locations. In terms of external situations, they both experienced opportunities and threats, such as increasing demand from third-world countries, digitalization, and the rise of niche/internet-based brands (McKinsey & Co. 2020). Due to these internal and external issues, both companies paid close attention to which organizational structure would best deal with these issues. Yet, each company showed that the duration for one organizational structure was longer than the other, resulting in an asymmetric shape. Our comparison case study was conducted to prove this asymmetry based on a case study method. Therefore, a discussion on the precedents regarding why P&G and Unilever showed different vacillation phases like the following is needed. The possible reasons for the research with each proposition are as follows.
Proposition 2. The initial structure that an organization adopted during its foundation will be more likely to become its dominant structure.

Proposition 2a.
Organizations that started with a centralized structure tend to stay in centralization longer than decentralization.

Proposition 2b.
Organizations that started with a decentralized structure tend to stay in decentralization longer than centralization.
Second, P&G and Unilever showed different vacillation patterns, depending on the geographic market (country) where most of their revenue was generated. This is closely related to the concept of psychic distance stimuli, which has long been recognized as a major impediment in MNEs' international expansion (Dow 2000;Hutzschenreuter et al. 2014). Psychic distance stimuli include major perceived factors of inter-country differences such as governance, cultural, and geographic distance. Okechukwu and Adiele (2019) showed that psychic distance indeed influences the effectiveness of international business outcomes, and in order to enhance their business performance, MNEs in the FMCG industry are recommended to thoroughly study the cultural, political and economic characteristics of foreign markets. When we consider this concept of psychic distance stimuli with asymmetric vacillation phenomenon, P&G has focused on the Western market, and more than half of its global revenue came from the United States and Europe, while Unilever successfully gained a market share in the developing-countries market. Therefore, considering both companies' host market and major consumer country targets, Unilever's psychic distance stimuli to its major market would be relatively higher than that of P&G. This finding can be also related to Oh and Contractor (2014), who revealed that most organizations can enhance their performance by focusing on their home regions rather than foreign regions.
The impact of the relationship between headquarter and local representatives on organizational outcomes also have been empirically proven by various researchers. In Solberg's (2002) study, the author revealed that the centralized power of headquarters facilitated more negative reactions at the intermediary level (Solberg 2002). On the other hand, when local managers tend to have more decision-making power and more autonomy, local representatives tended to have better relationships. Therefore, a European MNE such as Unilever seems to fit clearly into this definition of a decentralized structure, whereas the United-States-based P&G is likely located somewhat toward the border between the federation and confederation, which has fewer perceived cultural differences and less of a need to achieve customization.
Moreover, P&G started in the United States market, which was relatively homogeneous, and Unilever grew in the European market where the customers' needs, languages, and cultures were much more diverse than in a single market. For P&G, it did not have to accommodate its product packages or recipes, and thus the main strategic focus for the company was to centralize the overall manufacturing process so that it could reduce costs and focus more on R&D. Unlike P&G, Unilever had to consider various customers from diverse countries within the European market. Since the strategy of the firm tends to follow its organizational structure, what each organization's main organizational structure will be will shape both organizations' different strategies (Hall and Saias 1980).

Theoretical and Practical Implications
This case analysis demonstrated how two firms vacillated differently between two organizational structures in terms of the duration of each structure. It was found that there was an asymmetrical tendency for each firm to maintain one certain structure longer than the other while cycling between structures. P&G showed a longer duration of a centralized structure, and this pattern was mainly attributed to its origin as a centralized firm, enabling its success in the North American market. Also, due to changing initiatives for each globalization goal, P&G could stay in a decentralized structure through vacillation. It seems that after achieving economies of scale and cost reduction, globalization goals switch to goals of customization and geographic divergence, reducing the time spent in a decentralized structure. On the other hand, Unilever began its business under relatively decentralized circumstances with sprawling operating companies and facilities in Europe. This led to a longer duration of a decentralized structure. Several situational difficulties arose from standardization needs and cost reduction, and Unilever switched to a centralized structure. From our viewpoint from the literature review of the vacillation theory, a firm continuously undergoes managerial decision-making that deals with rising business problems, representing a swing shape between two different organization strategies with different widths of the swings. Based on the above results, this study suggests that a firm continues to vacillate between two different strategies to deal with the continuously changing business problems resulting from globalization, and how long a firm maintains one structure is asymmetrical.
Theoretically, this paper examined the key differences in organizational vacillation, especially focusing on MNEs in the fast-moving consumer goods industry. While prior studies on organizational vacillation theory mostly revealed what vacillation is and what the results of vacillation can be, how actual vacillation happens in a real-life case study has been missing. Thus, building upon vacillation theory, we depicted and analyzed how each firm's structure vacillated between centralization and decentralization. This study found that even though both firms vacillated alternately, as previous studies have already revealed, each stayed in the specific structure longer than the other. Our study focused on this difference and went further by categorizing each firm's strategies into centralization and decentralization and attempted to come up with a plausible explanation for this phenomenon. Moreover, unlike the existing studies that focused heavily on the impact of organizational ambidexterity and vacillation on firm performance, this paper focused on how vacillation itself happens with real business cases.
This study can contribute to organizational vacillation theory, since most of the existing studies presumed that organizations vacillate alternatively between different structures but did not focus on the specific duration of the vacillation for each firm. Asymmetric vacillation, which focuses on the 'time duration' of organizational vacillation, can have various effects on firm performance and ambidexterity. There is an existing study that examined the impact of vacillation 'frequency' and 'scale' (Kang et al. 2017); however, there may be yet another effect and explanation for the 'time duration' of organizational vacillation. Even though we adopted a case study method and did not empirically test the asymmetric vacillation, we believe that this study can provide guidance for further research.
Moreover, this paper also offered various explanations to understand why organizations vacillate asymmetrically rather than symmetrically. According to this paper, organizations might have their own dominant structure that can be associated with the market characteristics of the home country. This can be partially explained by the imprinting effect, which highlights the impact of the early environment on the organizational structure and processes. The existing studies regarding the imprinting effect revealed that organizations are more vulnerable to their external environment especially when they are in the early stages of their life cycle (Hannan et al. 1996;Stinchcombe 1965). Based on this effect, P&G was highly influenced by the United States market while Unilever was highly influenced by the European market, which are the home countries of each organization.
Practically, this study could provide the following implications to HRBPs (HR business partners) and HR partners of multinational enterprises and is not only limited to the fast-consuming goods industry. The discovery of this vacillation phenomenon and its characteristics could give useful insights to HR teams in terms of deciding whether to reshuffle or shake things up at the HQ and subsidiary level. As our study observed and demonstrated that the span of an organization in a certain structure can be affected by asymmetrical vacillation, paying attention to whether the organization will continue its structure, especially with a further demonstration of its influence on well-functioning exploitation and exploration, will be important. Moreover, our findings can provide significant implications for managers who aim to skillfully enhance organizational performance by vacillating between different organizational structures at the right time. Managers and management teams must understand how their own organizations vacillated in the past and what were the motivations and effects of those choices. Just as Nickerson and Zenger (2002) used the phrase 'sailing into the wind', leaders in organizations are much like a captain sailing into the wind who knows, understands, and predicts the waves.

Limitations and Future Research Directions
This study is not without limitations. Since this study was examined through a longitudinal analysis of certain periods of two firms, the findings of this study should be carefully generalized. Also, since the study heavily relied on content from media interviews and annual reports, a quantitative comparison was hard to achieve due to the lack of diverse materials and numerical data, including consistent product and country data. Thus, to further develop this study, research can be expanded to include generalizing the theory through statistical analysis and discovering variables that lead to asymmetry.
Based on these limitations, we recommend the following future research directions. First, future research can develop empirical measurements to test asymmetric organizational vacillation. After Nickerson and Zenger's (2002) study that theoretically argued how organizational vacillation may lead to ambidexterity, empirical tests of vacillation theory have been still missing (Kang et al. 2017). This study adopted a case study method to illustrate and examine specific patterns of asymmetric vacillation; however, a future study could expand on our results by empirically measuring asymmetrical vacillation with a large-scale sample. Even though we simply calculated the time duration for which the organizations stayed in what organizational structure, a future study could not only empirically measure asymmetric vacillation but also different speeds regarding the transition period. Second, future studies could expand organizational vacillation theory by adopting a qualitative research method. Our study relied heavily on secondary data sources such as annual reports and press releases; however, future studies could conduct interviews with top management teams and regional division leaders to investigate the actual effects of asymmetric vacillation by collecting primary data. Third, we believe that organizational research can be better informed by moving away from simple snapshots of strategies and, instead, exploring dynamics more. Rather than simply examining organizational ambidexterity or vacillation's effects on firm performance, future research could study in what conditions organizations tend to vacillate symmetrically or asymmetrically and could study whether under some conditions, adhering to one specific organizational structure may be a rational decision for a firm, just as Nickerson and Zenger's (2002) study revealed that inertia can yield significant efficiency-related benefits.