Who bears the costs of the UK soft drink tax? An empirical study of medium-term effects

Wiredu, Laura Article Who bears the costs of the UK soft drink tax? An empirical study of medium-term effects Junior Management Science (JUMS) Provided in Cooperation with: Junior Management Science e. V. Suggested Citation: Wiredu, Laura (2025) : Who bears the costs of the UK soft drink tax? An empirical study of medium-term effects, Junior Management Science (JUMS), ISSN 2942-1861, Junior Management Science e. V., Planegg, Vol. 10, Iss. 4, pp. 985-1008, https://doi.org/10.5282/jums/v10i4pp985-1008 This Version is available at: https://hdl.handle.net/10419/334185 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/4.0/ Junior Management Science 10(4) (2025) 985-1008 Junior Management Science www.jums.academy ISSN: 2942-1861 Editor: DOMINIK VAN AAKEN Advisory Editorial Board: FREDERIK AHLEMANN JAN-PHILIPP AHRENS THOMAS BAHLINGER MARKUS BECKMANN SULEIKA BORT ROLF BRÜHL KATRIN BURMEISTER-LAMP CATHERINE CLEOPHAS NILS CRASSELT BENEDIKT DOWNAR KERSTIN FEHRE MATTHIAS FINK DAVID FLORYSIAK GUNTHER FRIEDL MARTIN FRIESL FRANZ FUERST WOLFGANG GÜTTEL NINA KATRIN HANSEN ANNE KATARINA HEIDER CHRISTIAN HOFMANN SVEN HÖRNER STEPHAN KAISER NADINE KAMMERLANDER ALFRED KIESER ALEKSANDRA KLEIN NATALIA KLIEWER DODO ZU KNYPHAUSEN-AUFSESS SABINE T. KÖSZEGI ARJAN KOZICA CHRISTIAN KOZIOL MARTIN KREEB WERNER KUNZ HANS-ULRICH KÜPPER MICHAEL MEYER JÜRGEN MÜHLBACHER GORDON MÜLLER-SEITZ J. PETER MURMANN ANDREAS OSTERMAIER BURKHARD PEDELL ARTHUR POSCH MARCEL PROKOPCZUK TANJA RABL SASCHA RAITHEL NICOLE RATZINGER-SAKEL ASTRID REICHEL KATJA ROST THOMAS RUSSACK FLORIAN SAHLING MARKO SARSTEDT ANDREAS G. SCHERER STEFAN SCHMID UTE SCHMIEL CHRISTIAN SCHMITZ MARTIN SCHNEIDER MARKUS SCHOLZ LARS SCHWEIZER DAVID SEIDL THORSTEN SELLHORN STEFAN SEURING VIOLETTA SPLITTER ANDREAS SUCHANEK TILL TALAULICAR ANN TANK ANJA TUSCHKE MATTHIAS UHL CHRISTINE VALLASTER PATRICK VELTE CHRISTIAN VÖGTLIN BARBARA E. WEISSENBERGER ISABELL M. WELPE HANNES WINNER THOMAS WRONA THOMAS ZWICK Volume 10, Issue 4, December 2025 JUNIOR MANAGEMENT SCIENCE Marie-Claire Joyeaux,Work Less, Live More? The Impact of an Introduction of the Four-Day Working Week on Happiness in the Context of the Icelandic Four-Day Working Week Experiment Niklas Janßen, Integrating Sustainability in Risk Management and Internal Control Systems: An Empirical Assessment of ESG Reporting of German DAX40 Firms Tobias Keserü,A New Dimension of Transparency: ESG Disclosure and Its Effect on Shareholder Behavior Antonia Engel, ESG Regulation Across the Globe: Does ESG Regulation Pay Off? Finn Matthes Gooßen, The Impact of Female Board Members on ESG Performance: An Empirical Analysis Louisa Felicitas Bläßer, Implicit Measurement of the Moral Self-Image Using the Go/No-Go Association Task (GNAT) - An Empirical Investigation of the Convergent Validity Between Explicit and Implicit Measures Laura Wiredu, Who Bears the Costs of the UK Soft Drink Tax? An Empirical Study of Medium-Term Effects Thiemo Janßen, From Pictures to Perceptions: Exploring the Strategic Use of Visuals in CSR Reports and the Impact of Regulatory Mandates Vanessa Jeske, Determinants of Corporate Bond Mutual Fund Flows Felix Yumuşak, The Influence of Leadership Style on the Acceptance of Generative AI in the Workplace -The Role of Organizational Commitment, Job Insecurity and Interaction Frequency 831 858 876 904 940 966 985 1009 1028 1053 Published by Junior Management Science e.V. This is an Open Access article distributed under the terms of the CC-BY-4.0 (Attribution 4.0 International). Open Access funding provided by ZBW. ISSN: 2942-1861 Who Bears the Costs of the UK Soft Drink Tax? An Empirical Study of Medium-Term Effects Laura Wiredu Paderborn University Abstract Using five years post-tax data on CPI prices, national employment as well as firm-level employment, we provide novel evidence on the medium-term effects of the UK Soft Drinks Industry Levy (SDIL). Applying a difference-in-differences research design, we find that neither consumers nor employees visibly bore the costs of the tax. However, given the falling trend in beverage prices in the UK, consumers faced a 6% less decline in soft drink prices compared to prices of levy-exempt beverages, since the introduction of the tax. Employees in the soft drink manufacturing industry were unaffected by the tax, whilst employees in the beverage manufacturing industry even benefitted from it through significant employment increases. Our study offers policy advice for other countries which are yet to implement soft drink taxes. We advocate the implementation of tiered tax designs, such as in the UK, due to its positive impact on employment and minimal burden for consumers. Keywords: public health policy; soft drink taxation; UK sugar tax 1. Introduction Increasing health concerns, such as obesity and the associated non-communicable diseases (NCDs)1, have urged governments worldwide to address the excess sugar intake of the population (World Health Organization, 2023a). Hereby, the UK is one of the countries suffering the most from this health issue, having the third highest obesity rate in Europe in 2016, succeeding Malta and Turkey (Metcalfe & Sasse, 2023). By 2016, it was inevitable for the UK government to encounter this adiposity epidemic, with an obesity rate striking nearly 30% (Metcalfe & Sasse, 2023).2Consequently, the UK Soft I would like to express my sincere appreciation to my supervisor, Dr. Henning Giese, whose invaluable input and continuous support made this publication possible. My deepest gratitude also goes to my family, friends and (former) managers, without whom I would not be where I am today. Their unwavering support and trust gave me the courage to pursue and achieve my academic goals. 1Non-communicable diseases are chronic diseases, such as cardiovascular health issues (e.g. strokes), diabetes, cancer etc. They account for 74% of worldwide deaths (World Health Organization, 2023b). 2As shown in Figure 2 in Metcalfe and Sasse (2023). Drinks Industry Levy (SDIL)3was implemented in April 2018 (HM Revenue & Customs, 2024a). Its two-tiered policy design aims at encouraging manufacturers to reduce the sugar content of their beverage portfolio through reformulation (HM Revenue & Customs, 2024a); the less sugar is added to a drink the lower the tax burden. However, as prevalent for excise taxes, producers often shift the extra costs arising through the tax onto other stakeholder groups to maintain pre-tax profits. Considering that the UK soft drink tax is a producer-imposed tax, it is not intended to be passed down to consumers (e.g. in the form of higher product prices) or employees (e.g. job cuts). Thus, to understand the effectiveness of the UK soft drink tax, it is vital to first question “Who bears the costs of the UK soft drink tax?”. To identify which party carried the tax burden I perform differencein-differences analyses using firm-level and national employment data, as well as product price data on soft drink prices. I find that neither employees nor consumers visibly bear the tax burden. However, the tax introduction decelerated the 3For simplified comprehension, I hereafter speak of the UK soft drink tax in lieu of the UK Soft Drinks Industry Levy. DOI: https://doi.org/10.5282/jums/v10i4pp985-1008 © The Author(s) 2025. Published by Junior Management Science. This is an Open Access article distributed under the terms of the CC-BY-4.0 (Attribution 4.0 International). Open Access funding provided by ZBW. L. Wiredu /Junior Management Science 10(4) (2025) 985-1008986 general trend of falling soft drink prices in the UK. By using medium-term data, my study captures effects which have not been materialised in other previous short-term studies on employment and prices. This paper is structured as follows. Chapter 2outlines the institutional setting of the UK soft drink tax and offers an extensive literature review on soft drink taxation. In Chapter 3, I outline my empirical strategy and describe the data. Chapter 4elaborates on the empirical results. Robustness checks are presented in Chapter 5. Lastly, Chapter 6concludes, summarises limitations and offers an outlook on future research. 2. Theoretical Background 2.1. Institutional Setting Excise taxes have become a favourable tool to nudge the population’s consumption pattern into a certain direction. Often referred to as “sin taxes”, they are levied on the consumption of goods which are associated with (high) future costs for society (Allcott et al., 2019; de la Feria, 2024). For example, the intake of tobacco and alcohol or the use of fuel often come at the cost of the environment or a person’s health. By taxing these types of products, their prices increase, leading to a decline in demand and herewith limiting their negative impacts (de la Feria, 2024). Nowadays, excise taxes are often levied on sugar-sweetened beverages (SSBs),4in order to curb the population’s excess sugar intake. While the WHO’s recommendation on an adult’s daily sugar consumption is 50 g, a regular 600 ml soft drink bottle5already exceeds this threshold (World Bank, 2020). Hence, worldwide obesity levels have been continuously rising since more than four decades (Metcalfe & Sasse, 2023), making excise taxes on sugary drinks inevitable nowadays. According to the World Health Organization (2023a), more than 108 countries have adopted sugar-sweetened beverage taxes as at July 2022. Whilst the UK is one of the European countries most affected by high obesity rates (Metcalfe & Sasse, 2023), it is also one of the most recent countries in Europe to have adopted a tax on sugar-sweetened beverages (see World Bank, 2020, Appendix 1). Thus, the UK Soft Drinks Industry Levy, a national tax on sugar-sweetened beverages, was announced in March 2016 and implemented in April 2018 (HM Revenue & Customs, 2024a). The excise tax is imposed on soft drink producers and importers and comprises all types of soft drinks with more than 5 g of added sugar per 100 ml. Alcoholic drinks with more than 1.2% alcohol content, milk products, as well as fruit and vegetable juices are exempt from the levy (HM Revenue & Customs, 2024a). A tiered volume-based tax rate is applied to motivate producers to reduce the sugar content of their beverages by reformulating them. The two tiers are 4SSBs are drinks which contain free sugars (i.e. added sugars and sugars naturally inherent in fruit juices and concentrates, syrups and honey) (World Health Organization, 2023a). 5A 600 ml soft drink bottle approximately contains 64 g of sugar according to the World Bank’s (2020) example. defined as follows: 18 pence per litre for beverages with a sugar content between 5 g to 8 g per 100 ml; 24 pence per litre for beverages with more than 8 g sugar per 100 ml (HM Revenue & Customs, 2024a). In April 2023, the tax was extended to also encompass concentrates with more than 5 g sugar per 100 ml. All concentrates are taxed at the higher rate of 24 pence per litre (HM Revenue & Customs, 2024b). Whilst a common consensus has revealed that sugarsweetened beverage taxes can successfully reduce sugar consumption and herewith improve public health (World Bank, 2020), the estimated level of impact varies significantly across studies and countries. This can be partially attributed to the different policy designs and through which channels they are passed through. Whilst uniform taxes aim at cutting consumption levels through higher product prices for consumers, tiered tax designs focus on reducing overall sugar levels in beverages through product reformulation on the manufacturer’s side (World Health Organization, 2023a). Despite most countries applying uniform tax rates (e.g. a single tax rate on all sugary products), the UK government decided to introduce a tiered tax rate to decrease the sugar content in soft drinks instead of just targeting consumption through prices. Thus, its primary objective is to incentivise soft drink manufacturers to reformulate their products and to introduce low-sugar alternatives in their product portfolio in order to evade the tax (HM Revenue & Customs, 2016). Therefore, the UK government provided manufacturers with a two-year lead time to improve their sugar-sweetened beverage portfolio beforehand. Hence, the change should already take place on the manufacturing side instead of just transferring the tax burden onto the consumer. As obesity levels kept on increasing (Metcalfe & Sasse, 2023), despite the introduction of the soft drink tax, the British government was forced to further intervene in the population’s diet choices. Hence, in April 2022, the government implemented a compulsory calorie labelling law. All restaurants and other businesses (e.g. online deliveries) offering non-prepacked food (i.e. served meals) are obliged to display calorie information on their menu (Department of Health and Social Care, 2022). I assume that this law could affect employment levels in the soft drink industry in a negative way due to its signalling effect. Consumers might refrain from purchasing soft drinks in a restaurant or cafe, once they are reminded of the high amount of calories on the menu; the tax signals the customer how unhealthy the beverage is. Thus, SSB consumption might decrease in restaurants, leading to an overall lower demand for soft drink manufacturing. This continuous decline in demand could eventually lead to job losses in the respective industry. However, I argue that this law is unlikely to distort my results, as its introduction only overlaps with my data by one year (i.e. in 2023). I also perform a robustness check, in which the overlapping year is excluded, and find that my results remain robust (see Chapter 6). L. Wiredu /Junior Management Science 10(4) (2025) 985-1008 987 2.2. Literature Review Excise taxes entail additional costs for soft drink firms and could jeopardise their profits. Thus, firms often tend to pass the tax burden onto other stakeholder groups, such as employees (e.g. reductions in staffing) or consumers (e.g. increases in prices). Hereby, current literature extensively analyses price effects, as most sugar-sweetened beverage taxes are transmitted through this channel. Effects on employment remain an understudied area. Due to the novelty of most soft drink taxes, the vast majority of empirical studies are only able to capture the short-term effects of SSB taxation. Whilst this might be the case for price effects, I expect structural changes in employment to take longer to materialise. For example, Scarborough et al. (2020) analyse the shortterm effects on prices evoked by the UK soft drink tax. They conduct an interrupted time series analysis, using more than 210,000 observations on soft drink sugar content and prices. Primarily, they divide the sample into own branded products from the supermarket chain and famous branded products, as well as low-levy products, which have a sugar content between 5 g to 8 g per 100 ml and high-levy beverages with a sugar content of 8g per 100 ml or above. Firstly, they discover a reduction of the sugar content for both, “no name” and branded sugar-sweetened beverages. Secondly, they find that prices increased by 11.8 pence per litre for branded products from the higher levy tier, passing through approximately half of the tax. Surprisingly, the prices of the branded lower taxed beverages decreased by 17.4 pence per litre. For own branded drinks the opposite effect occurred. Prices of higher taxed drinks decreased by 63 pence per litre whereas prices of lower taxed products increased by 69 pence per litre (Scarborough et al., 2020). Clearer industry responses have been observed in the Mexican and Chilean sugar sweetened beverage markets, after the implementation and increases of their national soft drink taxes. Cuadrado et al. (2020) conduct a time series analysis to regress the increase of the Chilean soft drink tax in 2014 on soft drink prices. Their analysis is based on consumer price data two years post tax implementation. The authors estimate a 5.6% price increase of carbonated beverages, which is equivalent to a 140% pass-through of the tax. Prices of untaxed beverages, such as juices, were unaffected. Contrary to the UK soft drink tax, the Chilean SSB tax was a consumer-imposed tax, aiming to reduce the affordability of soft drinks and to stimulate healthier substitution effects. As of that, the increase in soft drink prices has been the main intention of the Chilean soft drink tax policy, which can therefore be viewed as fruitful. In the case of Mexico, Aguilar et al. (2021) also identify a significant positive effect of the national SSB tax on prices of taxed beverages. Using household purchase data from Kantar World Panel, they conduct an event study. The authors estimate a 9.7% price increase of taxed soft drinks and a pass through rate on prices of 80%. Herewith, both tax designs exceed the WHO’s recommendation on a minimal pass-through rate of 20% for SSB taxes to effectively reduce consumption (World Bank, 2020). Another major study stems from Stacey et al. (2019), who analyse price effects in the frame of the South African SSB tax. The authors use Consumer Price Index (CPI) data to perform a before and after comparison one year post tax. They find that prices of carbonated beverages increased significantly, holding a pass-through rate of approximately 70%. Similar evidence on prices is also provided for the Saudi Arabian sugar-sweetened beverage tax. Alsukait et al. (2020) also use CPI price data to derive the price effects one year after Saudi Arabia’s soft drink tax introduction. Their difference in differences analysis estimates significant price increases of 55% for carbonated beverages (which relates to a pass-through rate of 110%), compared to untaxed beverage prices. As mentioned above, most analyses focus on the consumer tax incidence of sugar sweetened beverage policies. However, excise taxes can also be shifted onto employees, by decreasing wages or reducing the number of employees to maintain pre-tax profits. Most commonly, opponents of the tax argue that the UK soft drink tax leads to major job losses (Lauber et al., 2022) or a reduction in working hours (Cawley & Frisvold, 2023). They assume that the decline in soft drink demand harms the firm’s economic performance and eventually leads to staff cuts (Hattersley et al., 2020). However, prevailing literature on employment does not validate these claims. One of the main studies in this field stems from Marinello, Leider, Pugach, and Powell (2021), which analyses the effect of SSB taxes on employment in the US state Philadelphia. The authors conduct a pre-post analysis, based on employment data from the Bureau of Labor Statistics. Overall, they do not find a significant effect of the tax on the number of employees in the private sector. Similarly, Díaz et al. (2023) perform an interrupted time series analysis, using employment data from the Ministry of Labour in Peru. As they explore the effects of the national soft drink tax, they do not observe significant impacts on wage and employment levels in the beverage manufacturing sector (Díaz et al., 2023). Another study focussing on employment changes in the beverage manufacturing sector originates from Guerrero-López et al. (2017). Using monthly survey data on the soft drink and non-alcoholic beverage manufacturing industries, they find that the Mexican soft drink tax had no significant effect on employment in the respective industries. Furthermore, Gonçalves et al. (2024) provide recent evidence on employment effects in the setting of Portugal. As the Portuguese soda tax was implemented in 2017 and partially increased in 2019, the authors employ an event study design to derive shortand medium-term effects of the tax. They retrieve wage and employment data from annual tax declarations, covering a period from 2017 to 2019. The authors find that neither wages nor employment levels in SSB companies were affected. Given the novelty of the UK soft drink tax, literature on its tax incidence is still quite scarce. While most studies on SSB taxes in other geographical settings imply price increases and no effect on employment, this might not be the case for the L. Wiredu /Junior Management Science 10(4) (2025) 985-1008988 UK due to its different policy design. Compared to other soft drink taxes, which aim at increasing prices to reduce consumption of sugary drinks, the British soft drink tax is not intended to be levied on consumers. For the tax to meet its primary target of reducing the population’s sugar intake and herewith obesity levels, the sugar levels in soft drinks must be substantially decreased. This is underlined by the paper of Briggs et al. (2017), who argue that product reformulation has the greatest effect on the population’s health in the UK. The authors conduct several modelling studies to quantify potential industry responses to the soft drink tax. They find that reducing the sugar content in drinks, by reformulating products, would have the greatest effect on public health. However, the effect partially diminishes if the industry decides to pass the tax onto the consumer, i.e. in the form of higher product prices (Briggs et al., 2017). Hence, to analyse the effectiveness of the UK soft drink tax, it is vital to first question “Who bears the costs of the UK soft drink tax?”. My research question herewith explicitly aims to identify whether the tax burden is borne by the company or shifted onto consumers or the firms’ employees.6The importance of identifying the incidence of the UK soft drink tax lies in the main objective of the tax. It was implemented to combat the UK’s health pandemic of obesity by reducing sugar levels in drinks and herewith the overall sugar intake of the population (HM Revenue & Customs, 2016). Thus, it is intended to be primarily levied on the manufacturer side, to stimulate reformulation efforts through its two-tiered design (Scarborough et al., 2020). If passed onto consumers, in the form of higher product prices, the policy is less effective (Briggs et al., 2017). As reformulation can be costly and time intensive, manufacturers might not be necessarily encouraged to reduce the sugar content of their products and have the option to take the simpler approach of passing the tax down to consumers. A change in consumer choice, however, to purchase less sugary drinks, seems more difficult. Not having the option in the first place of purchasing so much sugar in soft drinks, as overall sugar levels have been reduced, seems more promising. Thus, my approach focusses on identifying the incidence of the UK soft drink tax, as understanding industry reactions to the tax is essential to determine its efficiency (World Bank, 2020). Whilst a pass through of the tax onto consumers still has a positive, but reduced impact on the population’s health (Briggs et al., 2017), I argue that a pass through onto employees does not affect public health at all and only bears negative consequences for society (e.g. less employment, lower wages). I postulate that companies can react in one of the following three ways to the tax. The first and most common option is to pass the tax onto consumers, by raising product prices. Another option for soft drink manufacturers is to 6I assume that shareholders will not significantly be impacted by the UK soft drink tax. This is because they react extremely sensitive to dividend pay-cuts, wherefore firms usually avoid this measure (Heiden, 2002) and only use it as a last resort. Thus, I argue that it is very unlikely that the soft drink tax burden is a valid reason for soft drink companies to reduce dividend payouts. pass the tax burden onto their employees, by either reducing current staff or not increasing wages appropriately. This approach could be taken if firms want to avoid raising prices in order to remain competitive with other beverage manufacturers (e.g. juice manufacturers, which are not affected by the tax). Lastly, firms could avoid passing the tax burden onto other stakeholders due to market competitiveness and company reputation (e.g. reducing staff could stimulate negative publicity), but also circumvent the tax burden if they reformulate their soft drinks to fall below the sugar threshold. This response would be the UK government’s preferred reaction to the tax, because reformulation leads to overall lower sugar levels in soft drinks. Hence, to identify how the British soft drink manufacturing industry reacted to the tax and who now bears the consequences of it (i.e. the tax burden), I postulate the subsequent hypotheses: H1: The UK soft drink tax leads to a decrease in employment in the soft drink manufacturing industry. H2: The UK soft drink tax leads to an increase in prices of soft drinks. The first hypothesis aims to identify if consumers bear the cost of the tax due to rising soft drink prices. As elaborated above, the simplest approach for manufacturers to avoid the tax burden is to solely increase soft drink prices. In line with Cawley and Frisvold (2023), I argue that the pass through of the tax highly depends on the elasticity of demand and supply7in a perfectly competitive market. If demand is inelastic, as there are no alternative products to the taxed soft drinks (i.e. no replacement product) and consumers are highly dependent on the good, the industry can fully pass the tax burden onto consumers (100% price pass through). Thus, despite the price increases, demand for soft drink products will not decline due to the consumers’ dependency on that specific product. However, in the scenario of perfectly elastic demand, meaning there are many substitute products to the taxed soft drinks on the market, the consumer will opt for more cost-efficient alternatives, if prices of SSBs are increased. Thus, firms will refrain from raising prices and will fully absorb the tax burden (0% price pass through) or even offer price reductions, in order to stay competitive in the market. In line with Cawley and Frisvold (2023), I assume that the industry response will be somewhere in between the first scenario of fully passing the tax onto consumers and the second scenario of totally absorbing the tax burden. Similarly, the reaction of cutting staff (Hypothesis 2) seems intuitive, as increasing soft drink prices leads to a lower demand for the product. This leads to a decline in soft drink manufacturing and thus, eventually reduces the number of 7Elasticity of supply refers to the change of product supply due to a price increase. The change in the demand for a product (e.g. higher or lower consumption of it) due to price increases/decreases is known as the elasticity of demand (World Bank, 2020). L. Wiredu /Junior Management Science 10(4) (2025) 985-1008 989 jobs in the soft drink manufacturing industry (Hattersley et al., 2020). This has been a typical argument put forward by the opposition to global soft drink taxes (Hattersley et al., 2020), which I deem to be plausible for the UK soft drink tax too. Alternatively, a possible counter argument to my second hypothesis is that consumers might substitute the now more expensive soft drinks with untaxed beverages such as lowcalorie drinks or water (World Bank, 2020). Since soft drink manufacturers often produce a wide range of beverages (e.g. Coca-Cola also produces levy-exempt products such as water and juices), the substitution of products does not necessarily result in any employment losses. Rather, soft drink manufacturers might even require additional labour for reformulation and innovation of new low-sugar alternatives. As to my knowledge, there has been no other empirical study explicitly analysing the UK soft drink tax effects on employment. As the few studies on the UK soft drink tax price effects provide twofold evidence, my analysis offers additional insights in this field. By evaluating employment and price effects within a medium-term horizon, my study accounts for structural changes that take longer to manifest in the market (e.g. structural shifts in employment and long-term manifested pricing strategies). 3. Empirical Method 3.1. Model and Methodology To answer my research question of who bears the costs of the UK soft drink tax, I take three different approaches. First, I examine employment from a microeconomic scope, using firm-level data. Then I analyse the effect on a macroeconomic level, using national aggregate employment data of the beverage and soft drink manufacturing industries. Finally, I estimate price effects using product price data of soft drinks. I use two different employment approaches for more robust results. While the microeconomic data encompasses the most meaningful companies in the soft drink industry based on their market capitalisation and defined by The Refinitiv Business Data Classification, the macroeconomic employment analyses include randomly selected soft drink manufacturers regardless of their size. See Section 3.2 for a detailed description of the data, including its limitations. I apply the same econometric framework to all analyses, namely difference-in-differences (DiD). This research design is common in policy analysis, as it compares the different outcomes of a treatment and control (comparison) group over time (i.e. before and after a new policy intervention) (Marcus & Sant’Anna, 2021). For DiD to identify the actual causal effect of the policy, the following prerequisite must be fulfilled. The treatment group should have developed in a similar way as the control group if no policy (treatment) had been implemented. This is also referred to as the parallel trends assumption (Angrist & Pischke, 2009; Marcus & Sant’Anna, 2021). Put in other words, both groups must have the same pre-intervention employment/price trends. To test this assumption, I make graphical comparisons and conduct event studies. Event studies are commonly used for assessing the parallel trends assumption, as they offer a more granular view on the treatment effect (Marcus & Sant’Anna, 2021). Compared to DiD, which analyses the average change in outcomes (between treatment and control group) before and after the event, event studies look at the effect of the treatment at each specific period. To estimate the causal effect of the soft drink tax on employment and price levels, I perform high-dimensional fixed effects regressions for all three specifications. Fixed-effects models account for unobserved variables, which are constant over time but differ across groups and herewith may affect the endogenous variable (Angrist & Pischke, 2009; Wooldridge, 2021). For example, the location of a company could influence the availability of labour due to the infrastructure and economic conditions in that area. My fixed effects regression model controls for this possible variation within units. As stated by Lee (2016), the fixed effects estimator is the preferred estimator for difference in differences models. 3.1.1. Microeconomic Employment Analysis To test my first hypothesis, that is, whether the UK soft drink tax leads to a decline in employment in the UK soft drink manufacturing industry, I estimate the following model: Log(Emplo yees)i;t=β0+β1·Taxt +β2·Treatmenti +β3·(Taxt∗Treatmenti) +βk·Xi;t+Yi;t−1+γp+ϵi;t (1) My dependent variable, Log(Emplo yees)i;treflects the logarithmic number of employees in each company (i) per year (t). The variable Treatmentiis an indicator variable, taking the value one if a company is operating in the soft drink industry and zero if companies are alcoholic beverage producers. Taxtis a binary variable, which is equal to zero for periods before the soft drink tax came into effect (2013-2017) and equal to one for years after the tax (2018 – 2023). My key variable of interest is the interaction term Taxt∗Treatmenti. It is also referred to as the treatment indicator, which takes the value of one if a unit is treated and situated in a period after the tax intervention. Otherwise, it takes the value of zero. Its coefficient β3, also known as difference-in-differences estimator, captures the causal effect of the UK soft drink tax on employment. As I use a loglinear specification, the coefficient estimates the difference in employment changes between soft drink firms and alcohol producing firms before and after the tax intervention. To control for possible confounders, the Unemplo yment Rate, GDP and Total Population are captured in the term Xi;t. I believe that these common macroeconomic indicators are likely to influence current employment levels in the soft drink manufacturing industry. Furthermore, I also control for company fixed effects (γp), such as the company’s location for L. Wiredu /Junior Management Science 10(4) (2025) 985-1008990 example. All other potential confounders are encompassed in the error term ϵi;t. Furthermore, my model also accounts for autocorrelation by including the first lag of the dependent variable (Yi;t−1). Autocorrelation exists when the error term correlates across different time periods (Doran et al., 2017). In specific, when a variable can be defined by its previous year’s value, this is called an autoregressive process (Das, 2019).8Put in other words, I argue that the number of employees is strongly influenced by the number of employees in the preceding year.9 This could be caused by a company’s hiring patterns or its economic growth over the past years. As I use annual data, it is sufficient to only control for the first lag (Doran et al., 2017). Moreover, difference-in-differences models are often prone to cluster problems (Lee, 2016). This happens when error terms are correlated within a group (e.g. company) or across time periods (Imbens & Wooldridge, 2009). For example, companies often adapt their firm-specific hiring policies, which influences their number of employees over time. I control for this possible confounder by using clusterrobust standard errors,10 as suggested by Cameron and Miller (2015). In line with Gonçalves et al. (2024), I therefore cluster at the company level. 3.1.2. Macroeconomic Employment Analyses In addition to the microeconomic analysis (i.e. firmlevel employment), I perform two macroeconomic analyses (i.e. aggregate employment) for more robust results. I first regress the soft drink tax on employment in the soft drink manufacturing industry and afterwards on employment in the entire beverage manufacturing industry. Thus, I estimate the subsequent regression model to derive the macroeconomic employment effects evoked by the UK soft drink tax: Log(Emplo yees)i;t=β0+β1·Taxt +β2·Treatmenti +β3·(Taxt∗Treatmenti) +βk·Xi;t+Yi;t−1+ϕq+ϵi;t (2) The dependent variable Log(Emplo yees)i;treflects either the logarithmic number of employees in the beverage manufacturing industry or in the soft drink manufacturing industry. Equivalent to Equation (1), the term Xi;tcomprises the Unemplo yment Rate,GDP and Total Population. 8When the variable is only influenced by its last year’s value, this is referred to as an autoregressive process of order one or just AR(1) (Das, 2019). 9I perform the Wooldridge test for serial correlation and find that my model suffers from autocorrelation issues. Furthermore, I also perform the Breusch-Godfrey test for autocorrelation, which tests if autoregressive processes larger than AR(1) (first lag) are present in the model (Doran et al., 2017). The test statistic indicates that my model does not suffer from autocorrelation beyond the first autoregressive process (AR(1)). 10 Cluster-robust standard errors control for both, cluster problems and for heteroskedasticity (Cameron & Miller, 2015). Again, Yi;t−1denotes the lagged dependent variable11 and ϵi;tthe error term. I further control for industry-specific fixed effects (ϕq). Furthermore, I use heteroskedasticity-robust standard errors, but do not cluster at group levels due to an insufficient number of clusters. Not addressing possible cluster issues within the model could lead to an underestimation of standard errors (Lee, 2016). However, if the number of clusters is too small, this could also lead to underestimated standard errors (Cameron & Miller, 2015; Cawley & Frisvold, 2017). While Cameron and Miller (2015) postulate that more clusters lead to better statistical inference, their example underlines that fewer than 10 clusters result in extremely high rejection rates of the null hypothesis. I therefore refrain from using only two industry clusters. Thus, regression results should be treated with caution. 3.1.3. Price Analysis Using the regression model below, I examine if the UK soft drink tax leads to increases in soft drink prices (Hypothesis 2): Log(Price)i;t=β0+β1·Taxt+β2·Treatmenti +β3·(Taxt∗Treatmenti) +βk·Xi;t+ωr+δs+ϵi;t (3) My dependent variable Log(Price)i;tdenotes the logarithmic real (inflation-adjusted) price per product (i) per year (t). In the above model specification, the term Xi;tencompasses the control variables Unemplo yment Rate,GDP, Total Population and Volume. I argue that the first three variables can influence price levels, as these macroeconomic indicators reflect the health of the economy. If for example, there is an economic downturn, reflected in a high unemployment rate or low GDP, people tend to purchase less, leading to a decline in demand and finally decreasing prices. With the variable Volume I also control for the effect of the product size on prices, because small sized drinks tend to be more expensive than larger volume packs (World Health Organization, 2023a). Furthermore, I also control for region (ωr) and shop type (δs) fixed effects. All the other variables (Tax,Treatment,Tax ∗Treatment, error term) are identical to Equation (1). Additionally, I cluster standard errors at shop level, to control for correlated outcomes within the same shop. Unfortunately, I was not able to include a lagged price variable in this model due to the missing panel structure of my data (see subsequent Section 3.2.3). 3.2. Data For a unified comparison, I standardise time periods and nearly all control variables for all three analyses.12 All 11 In line with the previous microeconomic model (see Equation (1)), I perform the same tests for autocorrelation and thus, include a lagged dependent variable. 12 Besides the uniform control variables Unemplo yment Rate,GDP and Total Population I additionally control for Volume in the price analysis. L. Wiredu /Junior Management Science 10(4) (2025) 985-1008 991 datasets cover periods from 2013 to 2023, ensuring balanced preand post-tax periods. I obtained my control variables GDP and Total Population from the World Bank website. The variable GDP reflects the UK’s annual Gross Domestic Product per capita in current US $ (World Bank, 2023a). Total Population encompasses all residents in the UK independent of their legal status in the country (World Bank, 2023b). I obtain the annual Unemplo yment Rate from the Office for National Statistics. The variable depicts the proportion of unemployed individuals aged 16 years or older in relation to the working population (Office for National Statistics, 2020). Missing values of the dependent variables and controls are excluded from all analyses. Detailed summary statistics and descriptions of all variables can be found in the Annex in Tables A1 and A2. 3.2.1. Microeconomic Employment Data I obtain annual employee data from Refinitiv Eikon Worldscope, a global database which provides financial and industry information on public companies. Refinitiv is renowned for its trusted global equity indices used for market analysis. Their indices capture significant firms in a specific stock market, by including 99.5% of all publicly traded shares in a (trading) market (Refinitiv, 2023). Their sector indices are built upon the global equity indices, by appointing companies into pre-defined industry groups according to the TRBC. The Reference Data Business Classification (TRBC) is a categorisation system which groups companies in accordance with their business operations. The TRBC divides the beverage industry group into the following industries: brewers, distillers & wineries and non-alcoholic beverages. As the latter definition is quite broad, the database also offers more granular information on the company’s main economic activity with the NAICS code (North American Industry Classification System). I use the Refinitiv sector index of the UK beverage industry to derive a comprehensive constituent list of 15 UK beverage manufacturers. However, I restrict my analysis to companies having their main sales market in the UK, due to the subsequent reasons.13 As I hypothesize that soft drink companies shift the tax burden onto their employees, I am focussing on companies directly affected by the soft drink tax. Hence, I postulate that beverage manufacturers must have their main sales market in the UK, for a portion of their sales to be noticeably affected by the new tax burden. Herewith, I exclude three companies14 from the initial constituent list. 13 My definition of main sales market is the following: either the United Kingdom (UK) or Great Britain (GB) must be within a company’s three biggest sales markets according to their annual report in 2023. 14 I exclude East Imperial as the EU is only its fifth biggest sales market, accounting for 10% of their annual sales in 2022 (London Stock Exchange, 2023). Rogue Baron is excluded from the analysis, as no employment data was available in Worldscope. Momentous Holdings Corp was also excluded, as it is a holding company (not directly involved in beverage manufacturing itself), which currently does not engage in any business operations, has no revenue (Reuters, n.d.) and does not have any publicly available information. See Table 1for a detailed constituent list and information on the companies’ UK market shares. Table 1: Microeconomic Employment Data # Treatment Group Market Cap. in £b UK Sales in % 1Coca-Cola Europacific Partners 26.587 17.5% (GB) 2 Britvic 2.406 71,4% 3 Fever-Tree Drinks 1.193 32% 4 A G Barr 0.675 95.6% 5 Nichols 0.365 74% Control Group 1 Diageo 57.633 9.1 % (GB) 2 Chapel Down Group 0.114 97.8% 3 Daniel Thawties 0.044 100% 4 Gusbourne 0.039 78.8% 5Artisinal Spirits Company 0.027 29.8% 6 Adnams 0.006 98% 7 Distil 0.004 90.2% Note. Author’s own overview based on market capitalisation data retrieved from Refinitiv (2024a) and sales data retrieved from annual reports in 2023 of the respective companies. The graph depicts 12 of the 15 constituents of the TRBC beverage industry group, which are included in my regression analysis. Market capitalisation is depicted in billion pounds. Based on the TRBC of the UK beverage sector, I appoint companies to the treatment and control group as follows. Enterprises operating in the non-alcoholic beverage market are assigned to the treatment group,15 and hereafter referred to as soft drink companies, as their main economic activity lies in soft drink manufacturing according to their NAICS code. The control group contains companies from the distillers & wineries and brewers industries and are hereafter referred to as alcoholic beverage companies. My final sample consists of 12 beverage manufacturers, resulting in a total of 106 firm-year observations. I have an unbalanced panel, as not all companies reported employee data for each year. My dependent variable of interest, the number of employees, comprises both, full and part-time employees in a company (Refinitiv, 2024b). The number of employees range between seven to thirty thousand. In line with Guerrero-López et al. (2017), I use the logarithmic number of employees to control for the right skewness (no normal distribution) in the dependent variable. Due to the high variance in the number of employees, using the logarithm also makes 15 Contrary to the TRBC classification, I appoint Fever-Tree Drinks to the treatment group, as its main product segment are carbonated nonalcoholic mixers, which are liable to the UK soft drink tax (Young, 2016). L. Wiredu /Junior Management Science 10(4) (2025) 985-1008992 the final results more comparable, as they will be interpreted as percentage changes instead of absolute values. One limitation of my data is that the sector indices only cover publicly listed companies. However, Figure 1indicates that the UK beverage market is dominated by a few, very large manufacturers. The graph depicts the UK’s major beverage companies in terms of their market capitalisation as at October 2024 (Statista, 2024). It indicates that Diageo, Coca-Cola Europacific Partners and Britvic have a substantial lead in terms of market shares. I argue that manufacturers which are not listed on the stock exchange, are usually insignificant in size with a low number of employees, compared to the leading beverage manufacturers depicted in Figure 1. Furthermore, I partially account for these smaller manufacturers in my macroeconomic employment analyses (see Section 3.2.2). Hence, both microand macroeconomic employment analyses are representative for employment in the UK beverage sector. 3.2.2. Macroeconomic Employment Data To analyse the effect of the soft drink tax on macroeconomic employment, I retrieve annual employment data per industry from the Business Register and Employment Survey (BRES). This annual survey covers 85,000 randomly selected companies registered for VAT and PAYE16 in Great Britain and is considered as the governments predominant source for employment estimates (Office for National Statistics, 2022a). In the survey industries are categorised according to the UK Standard Industrial Classification (SIC). This system appoints companies into pre-defined industries based on their primary economic activity. SIC codes offer a five-level breakdown, from the section level (one digit) to the subclass level (five digits) (Office for National Statistics, 2009). I derive my statistics from the division level (e.g. manufacturer of beverages) and the class level (e.g. manufacturer of soft drinks, production of mineral waters and other bottled waters). Based on this classification system, I define my treatment and control groups as follows. For my first macroeconomic analysis, I use manufacturers of soft drinks, production of mineral waters and other bottled waters as a treatment group. My control group consists of six alcoholic beverage manufacturing industries, which I argue to be not significantly affected by the tax reform, as beverages with an alcohol content of 1.2% or more are exempt from the tax. For my second broader macroeconomic analysis, the beverage industry serves as the treatment and the textile industry as the control group. I argue that the textile industry is unlikely to be affected by the soft drink tax and therefore offers a robust comparison group. It also fulfils the parallel trends assumption (see Chapter 4), as it exhibits the same pre-tax trend as the beverage industry before the reform in 2018. A constituent list of the treatment and control groups for my macroeconomic analyses can be derived from Table 2. 16 Pay as You Earn (PAYE) is the UK income tax system, which directly deducts the tax from the employees’ wages (HM Revenue & Customs, n.d.). For the soft drink manufacturing analysis, I have a balanced dataset consisting of 22 industry-year observations over the period 2013 to 2023. The sample for the analysis of the beverage industry contains 71 industry-year observations and is an unbalanced panel. For both analyses, the dependent variable of interest is the logarithmic number of employees, which encompasses fulland part-time employees in an industry (Office for National Statistics, 2022a). A limitation of this study is that manufacturing data is only available for Great Britain and not the entire United Kingdom. A separate Business Register and Employment Survey is conducted for enterprises situated in Northern Ireland and partially provides employment figures per industry as of 2016. However, as this survey does not consistently report employment statistics for the sub-classes of the beverage industry (SIC codes 1101-1107), I exclude these statistics from my analysis and continue with using Great Britain’s employment data only. I postulate that my estimates are representative for the complete UK beverage manufacturing industry, as only a small portion of the manufacturers are situated in Northern Ireland (3.6%). Table 3underlines that my GB data represents a significant portion of the UK soft drink manufacturing industry (96.4%). Secondly, BRES employment data has an inconsistent definition. Data from 2009 to 2015 only surveys VAT-registered businesses in GB, whereas data as of 2016 onwards additionally includes PAYE businesses. Furthermore, I acknowledge the fact that my data does not distinguish between employees in soft drink, mineral water and bottled water manufacturing. 3.2.3. Price data I retrieve annual product price data (CPI price quotes in January of each year) from the Office for National Statistics (ONS) website (Office for National Statistics, 2024b). ONS publishes the constituent list of the Consumer Price Index (CPI), namely the CPI price quotes, on a monthly basis.17 The CPI includes approximately 700 different products, for which prices are collected from roughly 20,000 different shops across the UK. Hereby, the CPI depicts a representative sample of consumer goods. If consumption patterns or product availability in the shop changes, some products get replaced. Prices are hand collected from randomly selected shops (Office for National Statistics, 2023). Furthermore, information on the region, product volume, shop code (i.e. which store the price was collected from) and shop type (i.e. if the price was collected from chains with more than 10 shops or from independent stores) can also be obtained from the dataset. Moreover, current prices today (nominal price values) must be made comparable over the years due to rising inflation and changing purchasing power. To convert nominal prices into real (inflation-adjusted) prices, the nominal values are divided by an inflation adjustment factor, for example 17 The data is not badged as National Statistics and is only made available for research purposes (Office for National Statistics, 2024b). L. Wiredu /Junior Management Science 10(4) (2025) 985-1008 999 Figure 6: Parallel Trends - Macroeconomic Employment (Beverage Industry) Note. Author’s own overview based on employment data retrieved from Office for National Statistics (2015,2022b). Graph shows the average logarithmic number of employees in the beverage manufacturing industry (treatment group) and in the textile manufacturing industry (control group) from 2013 to 2023. Logarithmic number of employees is normalised to one in 2017 (i.e. one year before the tax implementation). try continuously rises over the entire observational period, compared to its value in the reference year (2017). Furthermore, the beverage (treatment) and textile industries (control) both have a similar positive pre-tax trend in employment, indicating strong parallel trends. Whilst the textile industry exhibits fluctuations over time, the beverage manufacturing industry demonstrates a continuous positive trend in employment, even after the tax implementation. The event study of the beverage and textile manufacturing industries is depicted in Figure 7. Whilst the coefficient estimates seem slightly statistically significant in the first two years of the observational period (i.e. 2013 and 2014), they become insignificant afterwards. Put differently, the textile and beverage manufacturing industries have comparable logarithmic employment levels after 2015 onwards, making them valid treatment and control groups. Again, coefficient estimates for 2018 onwards are also statistically insignificant, suggesting that the soft drink tax did not lead to a significant change in employment levels of the beverage industry, compared to the change in employment levels in the textile industry. The subsequent Table 6presents the regression estimates for both analyses. First, I regress the tax event on the logarithmic number of employees in the soft drink manufacturing industry. Results are depicted in Columns (1) and (2). I then estimate the effect of the soft drink tax on the entire beverage manufacturing industry, of which results are presented in Columns (3) and (4). Columns (1) and (3) represent the coefficient estimates of regressing the tax on employment in the respective industries, excluding all control variables and fixed effects. The regression estimates of my adjusted models, as described in Equations (1) and (2), are shown in Columns (2) and (4). My regression coefficient of interest, the difference-indifferences estimator, is positive when regressing the soft drink tax on logarithmic employment in the soft drink manufacturing industry in Column (1). This implies a positive change in employment in the soft drink industry, compared to the employment change in the control group. When including controls, this induced change in employment in the soft drink industry becomes negative, relative to the control group. However, both effects are statistically insignificant. Thus, my macroeconomic regression results of employment in the soft drink industry are in line with my insignificant microeconomic regression result. Therefore, I can reject Hypothesis 1, which postulates that the UK soft drink tax leads to a decrease in employment in the soft drink manufacturing industry. Put differently, my results emphasise that employees in the soft drink manufacturing industry did not bear the cost of the soft drink tax, as employment levels have not significantly decreased due to the tax implementation. A possible explanation could be that either firms might have absorbed the extra costs of the new soft drink tax burden or have passed the costs onto consumers (see Section 4.3 for price regression results). Alternatively, soft drink companies might have even bypassed the tax by reformulating their high sugary drinks just in time, before the tax came into effect. My results are in line with current literature on soft drink taxation and employment in other countries, such as Peru, Portugal and the US (Díaz et al., 2023; Gonçalves et al., 2024; Lawman et al., 2019). They all found no significant effect of soft drink taxes on employment in the soft drink manufacturing industry in their respective countries of analysis. Contrary, when regressing the soft drink tax on employ- L. Wiredu /Junior Management Science 10(4) (2025) 985-10081000 Figure 7: Event Study - Macroeconomic Employment (Beverage Industry) Note. Author’s own overview based on employment data retrieved from Office for National Statistics (2015,2022b). Graph shows the difference in logarithmic employment changes between the beverage manufacturing industry (treatment group) and textile manufacturing industry (control group). Confidence intervals are plotted at 95 percent level. The last years’ coefficient of the interaction term (Treatment * 2023) was omitted due to collinearity. Hence, only a coefficient estimate of zero is plotted without any confidence interval. ment in the entire beverage manufacturing industry, results become highly significant (see Columns (3) and (4)). The difference-in-differences estimator of the basic unadjusted regression is positive and statistically significant (see Column (3)). Precisely, the coefficient of 0.274, which has to be back transformed first,24 indicates that employment levels in the beverage manufacturing industry increased by 32% more than employment in the textile manufacturing industry. In Column (4) I see that this effect becomes even larger when controlling for confounding variables. Due to the UK soft drink tax, employment in the beverage manufacturing industry increased by 39%25 more than employment in the textile industry. I also find that the Unemplo yment Rate, GDP,Total Population and the employment numbers of the previous year do not significantly influence employment levels, neither in the beverage nor in the soft drink manufacturing sector. Contrary to my initial assumption that employment levels will decline in response to the soft drink tax (see Hypothesis 1), I observe that employees in the beverage industry even benefitted from the tax, as employment has significantly increased in this sector (see Columns (3) and (4)). To put these percentage changes into relation, absolute employment numbers before and after the tax implementation can be retrieved from Table 7. The absolute number of employees increased in both industries, the soft drink and the 24 Coefficient estimates of log-linear models are interpreted as percentage changes (Wooldridge, 2013). In this case, the percentage change is calculated as follows: (e0,274 −1)∗100 =31.52%. 25 Again, for comprehension purposes, the percentage change is calculated as follows: (e0,327 −1)∗100 =38.68%. overall beverage manufacturing industry, after the tax was implemented. However, when controlling for the general growing trends in the industries and other influential economic indicators (e.g. GDP) in my difference-in-differences analysis, only the employment increases in the beverage manufacturing industry are significant and attributable to the soft drink tax. I ascribe this trend to non-soft-drink manufacturing companies in the beverage market, which might have benefitted from spillover effects. For example, alcoholic beverage manufacturers often diversify their product range, also offering non-alcoholic alternatives. By anticipating the new tax burden of their competitors (soft drink manufacturing firms), they might have created substitute products (e.g. low-sugar non-alcoholic beverages). Thus, these types of firms might have hired new employees for product innovation tasks and herewith increasing overall employment levels in the beverage manufacturing industry significantly. Whilst other studies have found no significant employment effects (Díaz et al., 2023; Gonçalves et al., 2024; Lawman et al., 2019; Marinello, Leider, & Powell, 2021), I argue that the UK has a different setting. Compared to other countries with often uniform tax rates, the UK soft drink tax has a two-tiered tax design to encourage reformulation. I postulate that this tax policy increased overall employment levels in the beverage manufacturing industry, due to increased demand for employees to fulfil new product innovation tasks. Employment levels in the soft drink industry were not significantly affected. This indicates that they might have not required additional staffing for reformulation, as they might have solely passed on the tax to consumers. However, as ar- L. Wiredu /Junior Management Science 10(4) (2025) 985-1008 1001 Table 6: Employees Macroeconomic Regression Soft Drink Manufacturing Industry Beverage Manufacturing Industry Log(Employees) (1) (2) (3) (4) Treatment*Tax 0.613 -0.129 0.274∗∗∗ 0.327∗∗∗ (0.529) (0.106) (0.068) (0.076) Treatment 1.605∗∗∗ -0.336∗∗∗ (0.342) (0.046) Tax -0.320 0.360∗∗ -0.042 -0.089 (0.527) (0.171) (0.058) (0.095) Unemployment Rate 0.043 -0.044 (0.096) (0.034) GDP -0.000 0.000 (0.000) (0.000) Total Population 0.000 -0.000 (0.000) (0.000) Lag Log Employees 0.354 -0.253 (0.308) (0.237) Observations 73 71 22 22 Industry FE No Yes No Yes Adjusted R20.082 0.957 0.703 0.706 Note. Author’s own estimations. Coefficient estimates are presented separately for the dependent variable Log(Employees) in the soft drink manufacturing industry (Columns (1) and (2)) and in the beverage manufacturing industry (Columns (3) and (4)). Columns (1) and (3) depict the coefficient estimates of regressing the tax event on logarithmic number of employees. The fully specified (adjusted) models in Columns (2) and (4) additionally control for Unemplo yment Rate,GDP,Total Population and industry fixed effects. Standard errors are heteroskedasticity-robust but not clustered due to insufficient number of industry groups. Standard errors in parentheses ∗p<0.10, ∗∗ p<0.05, ∗∗∗ p<0.01. Table 7: Average Absolute Number of Employees (Macroeconomic) # Soft Drink Manufacturing Industry Alcoholic Beverage Manufacturing Industry Beverage Manufacturing Industry Textile Manufacturing Industry Before Tax 10,000 5,151 37,400 52,400 After Tax 13,333 5,719 47,167 50,333 Absolute Difference 3,333 568 9,767 -2,067 Change in % 33.33 11.03 26.11 -3.94 Note. Author’s own overview based on employment data retrieved from Office for National Statistics (2015,2022b). The table shows the average number of employees in absolute values before (2013-2017) and after the tax was implemented (2018-2023) for each industry separately. The change in the average number of employees is depicted in percent. gued above, I believe that alcoholic beverage manufacturers have utilised the soft drink tax introduction to expand their product portfolio. Thus, requiring additional employees and herewith increasing overall employment in the beverage sector. 4.3. Price Regression Analysis As employees in the soft drink manufacturing industry did not have to bear the costs of the soft drink tax, and employees in the beverage manufacturing industry even benefitted from it, I now analyse whether the tax burden was passed onto the consumer. Thus, I regress the UK soft drink tax on soft drink prices. The following two graphs highlight the strength of my price regression analysis. Figure 8shows the development of L. Wiredu /Junior Management Science 10(4) (2025) 985-10081002 Figure 8: Parallel Trends - Prices Note. Author’s own overview based on price data retrieved from Office for National Statistics (2024b). Graph shows the average logarithmic real price per litre of soft drinks (treatment group) and levy-exempt beverages (control group) from 2013 to 2023. The logarithmic real price per litre is normalised to one in 2017 (i.e. one year before the tax implementation). Figure 9: Event Study - Prices Note. Author’s own overview based on price data retrieved from Office for National Statistics (2024b). Graph shows the difference in logarithmic real price per litre changes between soft drinks (treatment group) and levy-exempt beverages (control group). Confidence intervals are plotted at 95 percent level. the average logarithmic real price per litre26 of soft drinks (red line) and levy-exempt beverages (e.g. juices and water) (blue dashed line), relative to the reference year 2017. The trend of soft drink prices seems to be relatively stable over the complete observational period and only showing a slight decline in the pre-tax period. However, the prices of levyexempt beverages constantly decline until 2019 (and hereafter fluctuate) at a much higher rate. 26 Hereafter prices is used as synonym. I therefore conduct an event study to further investigate the parallel trends assumption. Figure 9depicts the differential changes between the logarithmic real price per litre of levy-exempt beverages and soft drinks. As all coefficients are statistically insignificant, it can be assumed that there was no difference in price changes of levy-exempt beverages and soft drinks. This shows that the parallel trends assumption for treatment and control group holds. Overall, the event study also indicates that the soft drink tax probably does not have a statistically significant impact on the soft drink prices, L. Wiredu /Junior Management Science 10(4) (2025) 985-1008 1003 as the coefficients remain statistically insignificant after the tax came into effect. As logarithmic percentage interpretations can be twofold at first sight in difference-in-differences analyses, I also provide an overview of average absolute prices before and after the tax implementation in Table 8. The overview shows that soft drink prices as well as prices of levy-exempt beverages in general decreased over my entire observational period between 2013 to 2023. Whilst the average soft drink price was £3.85 in the pre-tax period, soft drinks on average only costed £3.17 post tax, a total decrease of 68 pence. The same trend is also observable for levy-exempt beverages. To further test whether and to what extent this price change is attributable to the introduction of the soft drink tax, or whether these declines are a general trend of economic downturn, I conduct a difference-in-differences analysis. Table 8: Average Absolute Prices # Soft Drink Prices Levy-Exempt Beverages Mean Price per Litre before tax (£ 2018) 3.85 1.34 Mean Price per Litre after tax (£ 2018) 3.17 1.09 Difference in Prices -0.68 -0.25 Change in % -17.7 -18.5 Note. Author’s own overview based on price data retrieved from Office for National Statistics (2024b). The table shows the mean real price per litre in absolute values before (2013-2017) and after the tax (2018-2023) was implemented for treatment (soft drinks) and control group (levy-exempt beverages) separately. The change in average price is depicted in percent. The results of my price regression analysis are reported in Table 9. I first regress the soft drink tax event solely on my dependent variable logarithmic real price per litre (see Column (1)). The difference-in-differences estimator indicates that the tax leads to a positive change in soft drink prices relative to the change in prices of levy exempt beverages. However, this effect is not statistically significant. When controlling for confounding variables, as well as for region and shop fixed effects, I estimate a statistically significant positive effect on price changes (see Column (2)). As I know that prices of soft drinks declined over the complete observational period, I revise the interpretation of the difference-indifferences estimator. As its coefficient depicts the difference in price changes between the treatment group compared to the control group, a positive coefficient either indicates that prices increased more or declined less in the treatment group than in the control group. In my case, I find that, despite the overall declining trend of beverage prices in the UK, the soft drink tax led to a 6%27 less decline in prices of soft drinks 27 Again, for comprehension purposes, the percentage change is calculated as follows: (e0,058 −1)∗100 =5.97%. compared to prices of levy-exempt drinks. Put differently, even though the tax did not lead to significant price increases, it slowed down the general falling trend of soft drink prices. It can be said that, although the consumer did not visibly bear the soft drink tax burden in the form of higher product prices, I argue that they were still constrained by the tax, as soft drink prices would have declined even faster without the tax. To quantify how much of the tax burden was passed onto the consumer, a pass through rate is estimated. However, as emphasised by Stacey et al. (2019), pass through rates can only be retrieved from regression analyses for uniform tax rates. For tiered tax rates, such as the UK soft drink tax, based on the sugar content and the applicable tax threshold, the pass-through rate could be calculated for each specific product. However, as my data does not provide nutritional information, I am not able to calculate the actual pass-through rate of the UK soft drink tax. However, I estimate the change in absolute prices, provided in Table 8, for comprehension purposes only. While Table 8shows that soft drink prices (-17.7%) declined slightly less than levy exempt beverage prices (-18.5%), this does not depict the true causal effect of the soft drink tax. My difference-in-differences analysis, which controls for other influences on prices, finds that soft drink prices declined on average 6% less due to the soft drink tax and relative to levy-exempt beverage prices. While declining prices seem counterintuitive at first, this trend could be caused by a general decline in demand for certain beverages, which might lead to significant promotions and herewith price decreases. Putting this into context, one of the few studies in this field, which also analyses the price effects of the UK soft drink tax, yields similar results. Using additional data on sugar content, Scarborough et al. (2020) differentiate between the two tax levies (i.e. the lower levy threshold has a tax rate of 18 pence per litre and the higher levy threshold of 24 pence per litre). They find that prices of soft drinks from the lower levy tier decreased by 10.7 pence, holding a tax pass-through rate of -59%. However, they also find that prices in the higher levy tier increased by 7.5 pence on average, equivalent to a pass-through rate of the tax onto consumers of 31%. They also identify an overall declining price trend for beverages in the control group, which costed 1.5 pence less 50 days post tax implementation. Overall, my price results indicate that the UK soft drink tax decelerated the falling trend of soft drink prices. Whilst the general decrease in beverage prices is another research topic for itself, I argue that the beverage market is quite competitive, with a few large manufacturers predominating the market (see Figure 1). Thus, retail prices have to be kept as low as possible (e.g. by offering price promotions), in order for beverage companies to remain competitive. In view of the UK soft drink tax introduction, manufacturers were not able to hold soft drink prices as competitive as the prices of other levy-exempt beverages. This could be attributable to the new tax burden, which soft drink manufacturers partially passed onto consumers, in the form of fewer price promotions, compared to what they would have offered consumers, if the tax L. Wiredu /Junior Management Science 10(4) (2025) 985-10081004 Table 9: Price Regression Log(Price) (1) (2) Treatment*Tax 0.036 0.058∗∗∗ (0.037) (0.022) Treatment 0.700∗∗∗ 0.535∗∗∗ (0.099) (0.043) Tax -0.223∗∗∗ -0.134∗∗∗ (0.020) (0.017) Unemployment Rate 0.063∗∗∗ (0.010) GDP 0.000∗∗∗ (0.000) Total Population 0.000∗∗ (0.000) Volume -0.994∗∗∗ (0.035) Region FE No Yes Shop Type FE No Yes Observations 31,957 31,957 Adjusted R20.142 0.587 Note. Author’s own estimations. Column (1) depicts the coefficient estimates of regressing the tax event on logarithmic real price per litre (i.e. constant 2018 pounds). The fully specified model in Column (2) additionally controls for U nemplo yment Rate,GDP,Total Population,Volume, as well as region and shop type (i.e. chains with more than 10 shops or independent stores) fixed effects. Standard errors are heteroskedasticity-robust and clustered at shop code level (i.e. unique shop identifier, indicating from which shop the product prices were collected from). Standard errors in parentheses ∗p<0.10, ∗∗ p <0.05, ∗∗∗ p<0.01. would have not been implemented (see declining trend in soft drink prices before the tax). An alternative explanation is that manufacturers did not have to pass the soft drink tax onto consumers, if they had reformulated their soft drinks early enough to bypass the levy. Scarborough et al. (2020) found that one year after the UK soft drink tax implementation, sugar in taxable soft drinks28 was reduced by 2.13 grams per 100 millilitres on average. They observe that one year after the tax, 30.7% of the soft drinks that were tax liable (i.e. had a sugar content of 5 g per 100 ml or more), moved below the 5 g sugar threshold. This underlines that, in line with the UK soft drink tax objective, manufacturers substantially decreased the sugar levels in their products. Overall, I suggest that further analyses on the price effects associated with the UK soft drink tax need to be conducted to ensure more transparent results. 5. Robustness Checks I test the validity of my price and employment regression results by performing a variety of adjustments to my main 28 Referred to as intervention drinks in Scarborough et al. (2020). model specifications (Equations (1)–(3)). Results can be derived from Tables A2 to A4 in the Annex. The tests are described in detail below. The first two checks are applied uniformly to all model specifications (Equations (1) - (3)). Robustness check three is applied uniformly to all employment analyses (Equations (1)-(2)). Hereafter, specific checks for price and employment analyses are conducted, adapting to the needs of the respective datasets. 1) Covid-19 Pandemic The Covid-19 pandemic significantly impacted economies worldwide, and herewith not only affected employment but also prices due to high inflation. Thus, I re-run all regression analyses, limiting my observational period from 2013 to 2020, to account for the Covid crisis as confounding error (in employment and price estimates). All results are robust to this change (see Column (1) in Tables A2 - A4). 2) Tax Announcement Effect As the UK soft drink tax was announced in 2016, two years before its actual implementation, soft drink manu- L. Wiredu /Junior Management Science 10(4) (2025) 985-1008 1005 facturers might have already adjusted their product prices or employment levels to compensate the anticipated future costs of the tax. This behaviour, also known as the Ashenfelter Dip, could distort regression results when the treatment group significantly changes their behaviour in the pre-tax period (Lee, 2016). Therefore, I use the tax announcement as regressor. I discover that only my macroeconomic soft drink analysis yields the same results as in the main specification. Results from my macroeconomic beverage industry analysis remain positive but become insignificant (see Column (7) in Table A3). Positive microeconomic employment estimates, which were insignificant before, become significant. This indicates that the announcement of the soft drink tax-induced increases in employment of soft drink companies, compared to alcoholic beverage companies. As this significant effect in the pre-tax period could distort my main microeconomic regression estimates (Table 5), results have to be treated with caution. My price estimates, which were significantly positive in the adjusted specification (Table 9), become insignificantly negative (see Column (2) in Table A4). Overall, this indicates that the announcement of the UK soft drink tax had no effect on soft drink prices. 3) Separate Estimations of Fixed Effects and Lagged Model Specifications Angrist and Pischke (2009) emphasise that fixed effects regressions together with lagged dependent variables must be applied with caution due to their different assumptions. Thus, as suggested by the authors, I estimate each model separately. I apply this robustness check exclusively to my employment analyses, as no lagged variables are included in my price analysis due to the missing panel component of the data. The macroeconomic results remain robust. In Table A4, the microeconomic results remain positive when the lagged dependent variable is excluded (see Column (3)) but become insignificantly negative when excluding fixed effects (see Column (4)). However, I argue that the results of my main model specification (see Columns (1) and (2) in Table 5) are likely to be valid due to the subsequent reason. Angrist and Pischke (2009) postulate that separate estimations of fixed effects and lagged dependent models serve as a bracket in which the true causal effect lies. When the lagged dependent model specification is correct, but a fixed effects model is incorrectly applied, this leads to an overestimation of the actual effect (upper bound). Vice versa, if a lagged dependent model is incorrectly applied instead of the correct fixed effects model, this leads to an underestimation of the causal effect (lower bound) (Angrist & Pischke, 2009). Thus, my fixed effects estimation (see Column (3) in Table A2) is the upper bound and the lagged dependent variable model the lower bound (see Column (4) in Table A2), in which the true causal effect of my main model specification lies (see Column (2) in Table 5). 4) Employment Analyses 2016 to 2023 Furthermore, I control for the change in methodology of the Business Register and Employment Survey, from which my macroeconomic employment data is obtained. Whilst data between 2009 to 2015 only encompasses VAT registered businesses in Great Britain, data from 2016 to 2023 additionally includes smaller firms with less than 20 employees (i.e. Pay As You Earn [PAYE]businesses) (Office for National Statistics, 2017). To control for this inconsistency, I exclude all observations before 2016 in my macroeconomic employment analyses and find that my results are still robust (see Columns (5) and (10) in Table A3). 5) UK Primary Sales Market My microeconomic employment analysis only includes businesses for which the UK is at least their third biggest sales market. I tighten this restriction in the robustness analysis to encompass only companies with the UK as primary sales market. I herewith exclude the largest multinationals, namely Coca-Cola Europacific Partners and Diageo. Again, my regression results are robust (see Column (5) in Table A2). 6) Missing Pre-Tax Data In my microeconomic employment analysis, I control for the missing pre-tax data of companies. Hence, I exclude Artisinal Spirits, Adnams, Gusbourne and Distil from my analysis, and find that my initial results are valid (see Column (6) in Table A2). For my price analysis I also exclude items with insufficient pre-tax observations. Thus, I exclude all cola drinks (item IDs 212011, 21025 and 21026), a bottle of mixer (item ID 212022) and a water bottle (item ID 21025). My initial results are also robust to this test (see Column (3) in Table A4). 7) “Pseudo” Panel Data Lastly, to control for potential autocorrelation in my repeated cross-sectional price dataset, I construct an artificial (“pseudo”) panel structure. I compute the average price of each item per period, resulting in 143 unique item-year observations. This robustness check yields similar results as my initial price regression (see Column (3) in Table A4). 6. Conclusion This paper analyses the medium-term effects of the UK soft drink tax on employment and price levels. I find that neither consumers nor employees had to bear the cost of the soft drink tax visibly. However, despite the overall falling trend in beverage prices, consumers of soft drinks were minimally constrained by the tax, as they experienced a 6% less decline in soft drink prices (relative to levy-exempt beverages) than they would have without the tax implementation. Whilst the L. Wiredu /Junior Management Science 10(4) (2025) 985-10081006 tax had no significant impact on employment in the soft drink industry, I observe significant employment increases in the entire beverage industry. I postulate that alcoholic beverage manufacturers increased their staffing for product innovation purposes (e.g. low sugary non-alcoholic drinks), to attract soft drink consumers with their more competitive prices than those of their tax burdened competitors. By using five years post-tax data, my study offers novel insights on the more manifested structural changes in the beverage sector, which has not been accounted in the numerous short-term studies before. I herewith provide evidence for policymakers and the UK government, showing that the soft drink tax did not come at the cost of employees and only minimally burdened consumers. Based on my results, the government can revise and adapt the policy to further incentivise reformulation and herewith unburden the consumer. My study is limited in the following ways. Due to missing nutritional information, I was not able to distinguish the effect of the soft drink tax on lower levy and higher levy tier drinks. Hence, I also could not quantify the pass-through rate of the tax. Thus, I was also unable to identify as to how much firms are burdened with the tax or bypass it by reducing sugar in soft drinks just below the threshold (i.e. bunching). Another limitation is that I could not account for autocorrelation in prices, hence price estimates have to be interpretated with caution. Lastly, due to the lack of an adequate geographical control group, I was not able to control for potential spillover effects of the tax (e.g. shifts of employment from the soft drink industry into alcoholic beverage manufacturing industry). As my study and the few other studies in this field diverge largely in estimations, I encourage further price analyses to arrive at a common understanding on the tax incidence of consumers. Also, further research on the general pricing structure in the UK beverage market needs to be conducted, to comprehend the unusual trend of the overall decline in beverage prices. References Aguilar, A., Gutierrez, E., & Seira, E. (2021). The effectiveness of sin food taxes: Evidence from Mexico. Journal of Health Economics,77, 102455. https://doi.org/10.1016/j.jhealeco.2021.102455 Allcott, H., Lockwood, B. B., & Taubinsky, D. (2019). Regressive Sin Taxes, with an Application to the Optimal Soda Tax. The Quarterly Journal of Economics,134(3), 1557–1626. https://doi.org/10.1093 /qje/qjz017 Alsukait, R., Wilde, P., Bleich, S., & Singh, G. (2020). Evaluating Saudi Arabia’s 50% carbonated drink excise tax: Changes in prices and volume sales. Economics & Human Biology,38(Article 100868). http s://doi.org/10.1016/j.ehb.2020.100868 Angrist, J. D., & Pischke, J.-S. (2009). Mostly Harmless Econometrics: An Empiricist’s Companion. Princeton University Press. https://doi.org /10.2307/j.ctvcm4j72 Bank of England. (n.d.). Daily spot exchange rates against US Dollars. Retrieved October 20, 2024, from https://www.bankofengland.co .uk/boeapps/database/Rates.asp?Travel=NIxAZx&into=USD Briggs, A. D. M., Mytton, O. T., Kehlbacher, A., Tiffin, R., Elhussein, A., Rayner, M., Jebb, S. A., Blakely, T., & Scarborough, P. (2017). Health Impact Assessment of the UK Soft Drinks Industry Levy: A Comparative Risk Assessment Modelling Study. Lancet Public Health,2(1), e15–e22. https://doi.org/10.1016/S2468-2667(1 6)30037-8 Cameron, A. C., & Miller, D. L. (2015). A Practitioner’s Guide to ClusterRobust Inference. Journal of Human Resources,50(2), 317–372. https://doi.org/10.3368/jhr.50.2.317 Cawley, J., & Frisvold, D. (2017). The Pass-Through of Taxes on SugarSweetened Beverages to Retail Prices: The Case of Berkeley, California: The Pass-Through of Taxes on Sugar-Sweetened Beverages. Journal of Policy Analysis and Management,36(2), 303–326. https://doi.org/10.1002/pam.21960 Cawley, J., & Frisvold, D. (2023). Review: Taxes on sugar-sweetened beverages: Political economy, and effects on prices, purchases, and consumption. Food Policy,117(Article 102441). https://doi.org /10.1016/j.foodpol.2023.102441 Cuadrado, C., Dunstan, J., Silva-Illanes, N., Mirelman, A., Nakamura, R., & Suhrcke, M. (2020). Effects of a sugar-sweetened beverage tax on prices and affordability of soft drinks in Chile: A time series analysis. Social Science & Medicine,245, 112708. https://doi.org /10.1016/j.socscimed.2019.112708 Das, P. (2019). Econometrics in Theory and Practice. Springer Singapore. htt ps://doi.org/10.1007/978-981-32-9019-8 de la Feria, R. (2024). Non-(Fully) Harmonised Excise Taxes and Irrebuttable Presumptions. EC Tax Review,33(3), 98–108. https://ssrn.com/a bstract=4775842 Department of Health and Social Care. (2022). New calorie labelling rules come into force to improve nation’s health. https://www.gov.uk /government/news/new-calorie-labelling-rules-come-into-forc e-to-improve-nations-health Díaz, J.-J., Sánchez, A., Diez-Canessco, F., Jaime Miranda, J., & Popkin, B. M. (2023). Employment and wage effects of sugar-sweetened beverage taxes and front-of-package warning label regulations on the food and beverage industry: Evidence from Peru. Food Policy,115, 102412. https://doi.org/10.1016/j.foodpol.2023.102412 Doran, J., Bourke, J., & Kirby, A. (2017). Introduction to Applied Econometrics Analysis Using Stata. Oak Tree Press. https://de.everand.com/bo ok/359276951/Introduction-to-Applied-Econometrics-AnalysisUsing-Stata Gonçalves, J., Merenda, R., & Pereira dos Santos, J. (2024). Not so sweet: impacts of a soda tax on producers. International Tax and Public Finance,31, 1388–1412. https://doi.org/10.1007/s10797-02309808-7 Guerrero-López, C., Molina, M., & Colchero, M. A. (2017). Employment changes associated with the introduction of taxes on sugarsweetened beverages and nonessential energy-dense food in Mexico. Preventive Medicine,105, 43–49. https://doi.org/10.10 16/j.ypmed.2017.09.001 Hattersley, L., Fuchs, A., Gonima, A., Silver, L., & Mandeville, K. (2020). Business, Employment, and Productivity Impacts of Sugar-Sweetened Beverages Taxes (tech. rep.). World Bank. https://hdl.handle.net /10986/34082 Heiden, S. (2002). Kursreaktionen auf Dividendenankündigungen: Ereignisstudie am deutschen Kapitalmarkt. Deutscher Universitätsverlag. https://doi.org/10.1007/978-3-322-89777-0 HM Revenue & Customs. (n.d.). Income Tax: introduction. Retrieved October 20, 2024, from https://www.gov.uk/income-tax/how-youpay-income-tax HM Revenue & Customs. (2016). Soft Drinks Industry Levy. Retrieved July 6, 2024, from https://www.gov.uk/government/publications/s oft-drinks-industry-levy/soft-drinks-industry-levy HM Revenue & Customs. (2024a). Soft Drinks Industry Levy statistics background and references. Retrieved November 6, 2024, from https ://www.gov.uk/government/statistics/soft-drinks-industry-lev y-statistics/soft-drinks-industry-levy-statistics-background-andreferences HM Revenue & Customs. (2024b). Soft Drinks Industry Levy statistics: Commentary 2024. Retrieved November 6, 2024, from https://www L. Wiredu /Junior Management Science 10(4) (2025) 985-1008 1007 .gov.uk/government/statistics/soft-drinks-industry-levy-statisti cs/soft-drinks-industry-levy-statistics-commentary-2021 Imbens, G. W., & Wooldridge, J. M. (2009). Recent Developments in the Econometrics of Program Evaluation. Journal of Economic Literature,47(1), 5–86. https://doi.org/10.1257/jel.47.1.5 Kumaranayake, L. (2000). The real and the nominal? Making inflationary adjustments to cost and other economic data. Health and Policy Planning,15(2), 230–234. https://doi.org/10.1093/heapol/15 .2.230 Lauber, K., Rippin, H., Wickramasinghe, K., & Gilmore, A. B. (2022). Corporate political activity in the context of sugar-sweetened beverage tax policy in the WHO European Region. European Journal of Public Health,32(5), 786–793. https://doi.org/10.1093/eurpub/ck ac117 Lawman, H. G., Bleich, S. N., Yan, J., LeVasseur, M. T., Mitra, N., & Roberto, C. A. (2019). Unemployment claims in Philadelphia one year after implementation of the sweetened beverage tax. PLoS ONE,14(3), e0213218. https://doi.org/10.1371/journal.pone.0213218 Lee, M. (2016). Matching, Regression Discontinuity, Difference in Differences, and Beyond. Oxford University Press. https://doi.org/10.1093/a cprof:oso/9780190258733.003.0005 London Stock Exchange. (2023). Audited full-year results. Retrieved October 2, 2024, from https://www.londonstockexchange.com/news-ar ticle/EISB/audited-full-year-results/15938249 Marcus, M., & Sant’Anna, P. H. C. (2021). The Role of Parallel Trends in Event Study Settings: An Application to Environmental Economics. Journal of the Association of Environmental and Resource Economists,8(2), 235–275. https://doi.org/10.1086/711509 Marinello, S., Leider, J., & Powell, L. M. (2021). Employment impacts of the San Francisco sugar-sweetened beverage tax 2 years after implementation. PLoS ONE,16(6), e0252094. https://doi.org/10.137 1/journal.pone.0252094 Marinello, S., Leider, J., Pugach, O., & Powell, L. M. (2021). The impact of the Philadelphia beverage tax on employment: A synthetic control analysis. Economics & Human Biology,40, 100939. https://doi.o rg/10.1016/j.ehb.2020.100939 Metcalfe, S., & Sasse, T. (2023). Tackling obesity (tech. rep.). Institute for Government. https://www.instituteforgovernment.org.uk/publi cation/tackling-obesity NISRA. (2023). BRES Publication and Tables 2022. BRES 2022 Detailed Tables – Excel Format. Retrieved October 28, 2024, from https://w ww.nisra.gov.uk/publications/bres-publication-and-tables-202 2 Office for National Statistics. (2009). UK Standard Industrial Classification of Economic Activities 2007 (SIC 2007): Structure and explanatory notes. https://www.ons.gov.uk/methodology/classification sandstandards/ukstandardindustrialclassificationofeconomicacti vities/uksic2007 Office for National Statistics. (2015). Business Register and Employment Survey (excluding units registered for PAYE only) : open access (2009-2015). Retrieved June 25, 2024, from https://www.nomi sweb.co.uk/query/construct/summary.asp?mode=construct&ve rsion=0&dataset=172 Office for National Statistics. (2016). Consumer Price Inflation Basket of Goods and Services: 2016. https://www.ons.gov.uk/econom y/inflationandpriceindices/articles/consumerpriceinflationbask etofgoodsandservices/2016 Office for National Statistics. (2017). Business Register and Employment Survey (BRES). Retrieved August 24, 2024, from https:// ww w.nomisweb.co.uk/articles/1062.aspx Office for National Statistics. (2018). CPIH Microdata Metadata Template. Retrieved November 10, 2024, from https://www.ons.gov.uk/e conomy/inflationandpriceindices/adhocs/008009cpihmicrodata metadatatemplate Office for National Statistics. (2020). A guide to labour market statistics. https://www.ons.gov.uk/employmentandlabourmarket/peoplei nwork/employmentandemployeetypes/methodologies/aguideto labourmarketstatistics Office for National Statistics. (2022a). Business Register and Employment Survey. Retrieved October 29, 2024, from https://www.ons.gov .uk/surveys/informationforbusinesses/businesssurveys/business registerandemploymentsurvey Office for National Statistics. (2022b). Business Register and Employment Survey : open access (2015 to 2022). Retrieved June 25, 2024, from https://www.nomisweb.co.uk/query/construct/summary .asp?mode=construct&version=0&dataset=189 Office for National Statistics. (2023). Consumer Price Inflation (includes all 3 indices – CPIH, CPI and RPI) QMI. https://www.ons.gov.uk/ec onomy/inflationandpriceindices/methodologies/consumerpricei nflationincludesall3indicescpihcpiandrpiqmi Office for National Statistics. (2024a). Consumer Price Inflation Basket of Goods and Services: 2019 - 2021 edition of this dataset Tables and annexes. https://www.ons.gov.uk/economy/inflationandpr iceindices/datasets/consumerpriceinflationbasketofgoodsandser vices Office for National Statistics. (2024b). Consumer price inflation item indices and price quotes. Retrieved June 26, 2024, from https : //www.ons.gov.uk/economy/inflationandpriceindices/datase ts/consumerpriceindicescpiandretailpricesindexrpiitemindicesan dpricequotes Office for National Statistics. (2024c). Unemployment rate (aged 16 and over, seasonally adjusted): %. Retrieved July 20, 2024, from http s://www.ons.gov.uk/employmentandlabourmarket/peoplenoti nwork/unemployment/timeseries/mgsx/lms Refinitiv. (2023). Refinitiv Global Equity Indices (tech. rep.). London Stock Exchange Group. https://www.lseg.com/content/dam/marketi ng/en_us/documents/methodology/global-equity-index-metho dology.pdf Refinitiv. (2024a). Constituents and Market Capitalisation of the UK Beverage Sector Index [Retrieved 12 June, 2024, from Refinitiv Eikon Terminal.]. Refinitiv. (2024b). Number of Employees Variable WC07011, 2013-2023 [Retrieved 12 June, 2024, from Refinitiv Eikon Terminal.]. Reuters. (n.d.). Momentous Holdings Corp (MMNT.PK). Retrieved October 26, 2024, from https://www.reuters.com/markets/companies /MMNT.PK/ Scarborough, P., Adhikari, V., Harrington, R. A., Elhussein, A., Briggs, A., Rayner, M., Adams, J., Cummins, S., Penney, T., & White, M. (2020). Impact of the announcement and implementation of the UK Soft Drinks Industry Levy on sugar content, price, product size and number of available soft drinks in the UK, 2015-19: A controlled interrupted time series analysis. PLoS Medicine,17(2), e1003025. https://doi.org/10.1371/journal.pmed.1003025 Stacey, N., Mudara, C., Wen Ng, S., van Walbeek, C., Hofman, K., & Edoka, I. (2019). Sugar-based beverage taxes and beverage prices: Evidence from South Africa’s Health Promotion Levy. Social Science & Medicine,238, 112465. https://doi.org/10.1016/j.socscimed .2019.112465 Statista. (2024). Leading beverage companies by market capitalization in the United Kingdom (UK) as of October 15, 2024. Retrieved October 31, 2024, from https://www.statista.com/statistics/1345274/u k-largest-beverage-companies-by-market-capitalization/ United Nations. (2018). Classification of Individual Consumption According to Purpose (COICOP) 2018 (Statistical Papers Series M No. 99). https://unstats.un.org/unsd/classifications/unsdclassifications /COICOP_2018_-_pre-edited_white_cover_version_-_2018-1226.pdf Wooldridge, J. M. (2013). Introductory Econometrics: A Modern Approach (5th). South-Western, Cengage Learning. Wooldridge, J. M. (2021). Two-Way Fixed Effects, the Two-Way Mundlak Regression, and Difference-in-Differences Estimators (Working Paper). https://doi.org/10.2139/ssrn.3906345 World Bank. (2020). Taxes on Sugar-Sweetened Beverages: International Evidence and Experiences (tech. rep.). https://openknowledge.worl dbank.org/entities/publication/9498c3ef-b425-54f1-aafd-d3d3 b8411674 World Bank. (2023a). GDP per capita (current US$) - United Kingdom. Retrieved July 25, 2024, from https://data.worldbank.org/indicat or/NY.GDP.PCAP.CD?locations=GB L. Wiredu /Junior Management Science 10(4) (2025) 985-10081008 World Bank. (2023b). Population, total - United Kingdom. Retrieved July 25, 2024, from https://data.worldbank.org/indicator/SP.POP.TOTL ?locations=GB World Health Organization. (2023a). Global report on the use of sugarsweetened beverage taxes 2023 (tech. rep.). https://www.who.in t/publications/i/item/9789240084995 World Health Organization. (2023b). Noncommunicable diseases. Retrieved July 20, 2024, from https://www.who.int/en/news-room/factsheets/detail/noncommunicable-diseases Young, S. (2016). Fever-Tree Aims to Mix It With Drinks Industry Elite. Reuters. Retrieved October 25, 2024, from https://www.reute rs.com/article/world/uk/fever-tree-aims-to-mix-it-with-drinksindustry-elite-idUSKCN0Y30WQ/