Plunging and bouncing back – the corona pandemic and business continuity of Finland’s seaports Open Access

The COVID-19 pandemic disrupted supply chains around the world, both local and global. When the World Health Organization (WHO) declared the pandemic in January 2020, significant declines in transport volumes and business revenues were observed from year 2020 until 2021, when recovery began to take place (OECD, 2022). The maritime sector as a whole and seaborne transports were among the most severely affected. Several studies have reported these impacts, including. Prathvi et al. (2021) on ports in India; Gu et al. (2023) on major Asian ports and Gu et al. (2024) on ports of Shanghai, Southampton, New York and Los Angeles; Wang et al. (2022) on Chinese port traffic; Cullinane and Haralambides (2021) on container operations in Rotterdam, Shanghai and Los Angeles and Chua et al. (2022) on the impacts and prospective counter-strategies of the shipping and offshore sectors, among others. However, the pandemic did not halt ports from developing their business and operations – for example, a major merger of port companies took place during the pandemic in Indonesia (Wiradanti et al., 2022). Notteboom et al. (2021) described the impact of COVID-19 pandemic on global container shipping and ports/terminal operators, specifically comparing the impact with the 2008–2009 financial crisis. Their analysis covered impacts on global supply chains, impacts on operational aspects, market structure and strategic behavior of shipping lines and terminal operators. They also addressed impacts on port activity levels, such as vessel calls, container volumes handled, and changes in container port connectivity. Most existing studies have been mainly observational and have rarely provided detailed measurements at the company or legal-entity level, instead focusing on overall sectoral or sub-sectoral impacts.

As businesses in the transport and supply chain industry clearly suffered from the pandemic, questions concerning resilience and business continuity have become increasingly urgent. Resilience in supply chains and transports have been widely studied, and the pandemic has emphasized the urgency of this research. There are obvious semantic overlaps and links between supply chain risk managements and supply chain resilience and vulnerability (Jüttner and Maklan, 2011). Prior to the pandemic outbreak, very few studies had thoroughly examined how supply chains would be affected by such adversity. Existing studies mainly addressed overall resilience strategies and frameworks (e.g. Gunasekaran et al., 2015, Pettit et al., 2013), role of technologies (e.g. Min, 2019) and collaboration (e.g. Scholten and Schilder, 2015), as well as conceptual development of resilience (e.g. Tukamuhabwa et al., 2015). However, in their literature review, Hohenstein et al. (2015) concluded that “… review reveals a strong need for an overarching SCRES [Supply Chain RESilience] definition and a clear terminology for its building elements. It indicates that most research has been qualitative and lacks in assessing and measuring SCRES performance”.

The United Nations Conference on Trade and Development (UNCTAD) (2022) concluded that many pre-pandemic issues were overshadowed by the pandemic such as cybersecurity concerns, infrastructure deficiencies and capacity constraints. Based on their global survey, UNCTAD reported that many stakeholders emphasized resilience management measures and investments as key lessons from the COVID-19 pandemic. These measures include disruption management practices, recovery planning for business continuity, the adoption for enhanced digital tools as well as improved management of critical human resources. Business continuity planning was one of the priority actions.

In its Guidebook on Resilient Maritime Logistics, UNCTAD (2023) has developed the port risk management and resilience-building toolbox. This toolbox consists of a set of tools and methods designed to strengthen port resilience to be implemented before, during and after a disruptive disaster, including business continuity management (BCM). BCM in the toolbox aims to identify risks threatening ports, analyze their potential consequences and support preparedness for and recovery efforts when disruptive incidents occur (UNCTAD, 2023). A successful BCM program enables ports to respond to disruptions, safeguard strategic objectives, protect their stakeholders’ interests and reputation and sustain value-creating activities program (UNCTAD, 2023). Critical activities in the BCM include conducting business impact analysis (BIA) and development of business continuity plan (BCP) (UNCTAD, 2023; Ono et al., 2016; Benavente et al., 2016).

As ports become increasingly integrated into supply chains, Loh and Thai (2016) developed a port-related supply chain disruption (PSCD) model. Validated and demonstrated, the model has shown that managing PSCD contributes positively to the identification of internal and external opportunities, strengthen the resilience of port operations and positively influences port’s financial health and market reputation. Vanlaer et al. (2022) studied ports’ resilience from an organizational perspective, viewing ports as critical infrastructures. Moreover, UNCTAD advises ports to have at least one single centralized BCP for the entire port or alternatively, more targeted BCPs for critical business activities or processes. Smaller ports are likely to implement a single BCP, whereas larger ports with multiple distinct and geographically dispersed units and operations should develop dedicated BCPs for each unit (UNCTAD, 2023).

The Great East Japan Earthquake caused severe damages to port facilities in the eastern Japan, resulting in major supply chain disruptions and significant economic losses (Ono et al., 2016). Similarly, Chilean ports are also prone to natural disasters. In the aftermath of Great East Japan Earthquake in 2011, Japanese and Chilean governments, universities and research institutions collaborated to develop an earthquake-and-tsunami-resilient society. One outcome was the creation of a methodological framework for preparing BCM in ports (Ono et al., 2016).

When considering business continuity ISO 22301:2019 (ISO, 2019) provides a natural starting point and baseline definition. The ISO standard outlines processes and practices how to ensure business continuity during exceptional sudden crises and shocks: how to prepare for the unexpected, maintain business operations continuity at an acceptable level during or immediately after a disruption and recover to a normal capacity. Research on resilience transport systems, along with more in-depth definitions, can be found from multiple references, for example from Leviäkangas and Michaelides (2014) and Leviäkangas and Aapaoja (2015). While the theoretical foundation on resilience in engineering or economics has not yet been established, some of the early definitions emerged from psychology and medicine. One of the most straightforward definitions states that “resilience means stable trajectory of healthy functioning after a highly adverse event” (Southwick et al., 2014). In psychology, resilience theory was outlined by Garmezy (1991). Additional references include Kahan et al. (2009), Norris et al. (2008) and Fiksel (2006). These early definitions clearly resonate with the idea of business continuity – in essence, resilience refers to the ability to maintain performance and functions in exceptional situations and crises, while being prepared for and proactive in anticipation for new crises. Most studies also emphasize the importance of ensuring “resilience resources”, which may be material (money, infrastructure, physical protection, etc.) or immaterial (information, cognitive capabilities, skills, etc.). The importance of managerial process of proactive (anticipation, identification, preparedness), concurrent (adaptation, absorption) and reactive (response, reaction) measures are recognized and highlighted consistently across the literature. BCM can therefore be regarded as a prescription of that managerial process. However, as Hillman and Guenther (2020) pointed out, much work remains to clarify the concept of resilience in terms of definitions, measurement and contextual positioning.

Historically, business continuity management (BCM) originated as a part of crisis management and further evolved in the 1970s (Herbane, 2010). Hiles (2014) described the development of BCM. In 1988, Disaster Recovery Institute International was founded in the USA while Survive! was established in the United Kingdom, promoting disaster recovery and eventually leading to the introduction of the Business Continuity Institute (BCI). A series of events such as terrorist attacks (e.g. US 9/11), fires and natural disasters (e.g. earthquakes, hurricanes, tsunamis) and growing critical dependence on information and communication infrastructures, increased awareness on the importance of physical security and resilient supply chains. As an example, the fire accident at Philips microchip plant in Albuquerque, New Mexico, in 2000 disrupted both Nokia and Ericsson. However, Nokia had a flexible supply chain which enabled them to resume production and launch new products (Hiles, 2014). Organizations then began to develop a more comprehensive approach, creating new roles beyond the traditional BCM, operational risk managers or crisis managers. These new roles are responsible for identifying and protecting the enterprise from adversities that threaten its viability (Hiles, 2014).

Herbane et al. (2004) defines BCM as a process that identifies an organization’s exposure to internal and external threats and synthesizes hard and soft assets to provide effective prevention and recovery. Moreover, BCM plays an integrated and strategic role that could harm an enterprise’s reputation. Organizations that recover quickly from crises are perceived by stakeholders as capable in managing recovery process effectively, are better in maintaining their competitive position (Herbane et al., 2004).

This study is motivated by the fact that over 90% of Finland’s exports and more than 70% of its imports are transported via seaports, measured in tonnes (Statistics Finland, 2025). When measured in value, ports handle more than 50% of the country’s foreign trade. Therefore, ports represent a clear lifeline for Finnish industry and economy in the exchange of goods. Ports are part of the country’s critical infrastructures, and are vulnerable not only to pandemics, but also to several other risks. For example, Yliskylä-Peuralahti et al. (2011) studied the impacts of stevedoring strikes in 2010.

The objective of this paper is to examine how business continuity of Finland’s seaport companies was affected by the pandemic. The motivation is justified by the findings of numerous studies examining the impact of pandemics on transport supply chains and the maritime sector, where ports are an essential element of both sectors. While most studies have been observational and focused on macro-level impacts based on a limited number of cases, this study draws a comprehensive dataset covering most seaport companies from a single country. Identifying which types of ports suffered the most or the least from the pandemic crisis provides a novel comparative perspective. Finnish seaports have previously been analyzed by Leviäkangas et al. (2011, 2015). However, no study with this breadth and coverage has been peer-reviewed and published, except for the work of Hilmola (2022) that modeled some of the pandemic effects on Finnish ports’ transport volumes, and Leviäkangas et al. (2023) and Leviäkangas (2024), who analyzed the financial impacts of COVID-19. The latter studies focused on financial key ratios and how they were affected by the pandemic. This paper instead focuses particularly on the business continuity aspect, analyzing not only the immediate impacts of the pandemic, but also how seaport companies managed to recover from the decline in turnover. Hence, this paper fills the gap in explicit resilience observations and measurements within a single-country context and covers an important role of the seaports as a critical part of the supply chain. Credible business continuity studies on seaports are scarce, with only a few studies (e.g. Loh and Thai, 2016, 2015, 2014). These studies discussed risks, port resiliency and resilience management at the general level. Pandemics, however, were not explicitly identified as risks, hence indirectly reflecting the unexpected nature of COVID-19 disruption.

Constituting resilience ex ante is conceptually and operationally demanding. For example, Bruckler et al. (2024) conducted a literature review and summarized 17 different types of indicators covering the phases of resilience curve (the curve presented later in Figure 1). This paper, by contrast, focuses on impact measurement in a straightforward manner that avoids overly complex conceptual or methodological approaches and does not contradict more sophisticated measurement techniques. Therefore, the results contribute to practitioners – by suggesting how to direct investments in resilience – and to academics tasked with defining and developing the concept of resilience and its application, as pointed out by Hohenstein et al. (2015). The complexity of resilience management has been universally acknowledged, not only in the supply chain context. While resilience management may be conceptually challenging, ex post observations of resilience can be made explicitly and without serious bias, as demonstrated in this study. These observations may provide the starting point for identifying better practices, managerial strategies and policies for enhancing resilience.

To summarize, the research question of this study is: how was business continuity maintained by different types of seaport companies during and after the pandemic shock? The typification parameters for port companies include company size, ownership, business focus and capital structure. It is intuitively appealing to assume that larger port companies, for example, would be more capable of surviving shocks than smaller ones. It is also reasonable to assume that different line of businesses and financial resources would play a role in withstanding adversities and recovering from them. The purpose of this paper is to investigate whether these parameters had any observable effect on the economic performance measured by turnovers. Two measurement indicators were used: turnover decline due to the pandemic and the net recovery of turnovers before and after the pandemic. It should be emphasized that observations presented here do not explain the effects themselves, rather as a first step in the empirical process (Leviäkangas et al., 2023).

The analysis covers seaport companies that have published their financial accounts in accordance with national corporate legislation and official accounting standards for limited liability companies (Finlex, 1997). Port companies have been required to do so since the corporatization of the Finnish port sector in 2014–2015 (Leviäkangas et al., 2023). The selected port company sample is explained in more detail in the following section. The focus remains on clearly measurable impacts derived from financial statements.

Since this study is based on official accounting data and past observations, it is strictly empirical and follows a uniform measurement system, namely financial reports that have legal document status. Empirical research is based on repeatable measurements and observations, where conclusions are based on those measurements and observations, and interpretation is carried out afterwards when the results have been analyzed using valid scientific methods. Accounting data can be considered uniform (identical measurements for all ports), reliable (legally binding documents) and repeatable, as it is produced in the same way for every accounting period. Should the accounting rules change, they change uniformly across all businesses within the same industry. However, only future experience can confirm whether the observations are valid in the long run or merely contextual and time-dependent and whether any solid theory can be derived from them. The approach follows the principles of grounded theory: no pre-determined hypotheses stated a priori (see, e.g. Chun Tie et al., 2019). Some observations are tested statistically during the process if allowed by the data. The idea is to discover first, and only then to assess whether there is something to be hypothesized.

From a methodological perspective, this empirical research combines both quantitative (accounting data) and qualitative (categorization of data) measurements. The latter are limited by our own judgments in creating meaningful categories, while the former depend on the accuracy and descriptive quality of accounting measurements. Descriptive statistics are employed to form assumptions and hypotheses; however, despite the limited available data, statistical testing and regression modeling are possible.

The research process was conducted according to the following steps, which also reflects the structure of this paper:

1. A brief review of the pandemic’s effects on seaports and maritime transport is conducted and the gap concerning business continuity measurements is identified (see Section 1, “Introduction”).

2. In Section 2, “Objectives and scope”, the main research question and other objectives of the study are elaborated.

3. In Section 3, “Methods and Data”, the data and measurements are explained in detail. Section 3.1 “Methodology”, provides an overview of the general methodology and approach. Section 3.2 presents the straightforward method for measuring the decline and recovery (measuring resilience). This method compares the change in turnovers with subsequent recovery. Seaports defined as independent legal units (i.e. registered limited companies) were included in the analysis and their financial statements were obtained from public registers. The financial accounting data are described in Section 3.3 (Port data), which forms the basis of the business continuity measurement. Seaports were also characterized in terms of attributes, size, line of business, ownership, level of specialization and type and direction of freight. These attributes are linked to business continuity measurements (see Section 3.3 Port data).

4. Analyses were first conducted using descriptive statistics and where these suggested further testing, statistical significance was assessed. The results are presented in Section 4 “Results”, with testing procedures detailed in Section 3.5 “Statistical testing”.

5. The results are discussed, evaluated and conclusions drawn on the implications of the study. Both practical and theoretical considerations are presented in Section 5 “Conclusion”.

In essence, the research process followed the classical tradition of exploratory empirical approach: starting with the definition of the problem or gap in the knowledge, followed by data collection and categorization, measurements and analysis and finally conclusions and interpretation. Some theoretical implications are also discussed in the final section.

Resilience and business continuity measurement were measured according to the logic illustrated in Figure 1. Several critical measurement points are considered:

1. When the system experiences a decline in performance due to the shock, yet it remains operational with a performance level that is lower than normal (p_N – p_L).

2. When the system’s performance is significantly reduced (p_R), and when it ultimately collapses (p_C), being unable to operate.

3. When the system recovers to operational (p_L) and eventually returns to normal (p_N) performance levels.

Time factor is also critical as it determines how quickly the system can recover to an operational performance level. In this study, the focus is on measuring the difference between normal level of performance and reduced level of performance. Despite the a priori assumption that there are differences between the ports as to how deeply they experienced the crisis and how quickly they recover, none of the measured objectives (the seaport companies) experienced any business continuity collapse.

The measurements are as follows: first, the performance level was measured as turnover changes. The change from 2019 to 2020 was decisive, as WHO declared the pandemic globally in January 2020. Second, the bounce-back and recovery was measured as turnover change between 2019 and 2021. If recovery was strong, the sum of the changes would be close to zero. If the recovery was incomplete by the end of 2021, then the sum of the changes would remain negative. The magnitude of the changes is also important: it is assumed that some port companies experienced a deeper decline in turnover than others.

As to the analysis of measurements, i.e. accounting data, mainly descriptive statistics with visualization were used. Some statistical tests of averages of accounting figures were conducted using pairwise comparisons.

Conceptually, the measuring of business continuity was carried out as follows:

{Turnover change between 2019 and 2020} was the measure for turnover plunge, and {Turnover change between 2019 and 2021} was the measure of recovery with the changes expressed in percentage units. Inflation was not considered to have any significant effect, since the recorded official consumer price increases were recorded at only 0.29% for 2019–2020 and 2.19% for 2020–2021 (StatFin, 2023). Moreover, as 2019 was used as the benchmark year for both the decline (2019–2020) and the recovery (2020–2021), the inflation effect would be irrelevant in comparative analysis, as the effect is uniform across the data set. Figure 2 illustrates the straightforward measurement system.

The research sample consisted of 18 seaport companies along the coast of mainland Finland. Table 1 lists the ports, including data on transport volumes, ownership, size, primary line-of-business, specialization, freight flow direction and main types of freight. Only legal port entities (registered limited liability companies) were included. A few large ports that were part of a large industrial entity, such as steel mills or petroleum refineries companies, were excluded from the dataset as they did not disclose public financial statements. The data coverage was quite extensive, representing more than 80% of the goods transport tonnage and passenger numbers. In total, 23 operational seaports on the mainland coastline in 2023 handled export and import cargo (Finnish Ports Association, 2024). International passenger traffic was fully represented by the sample except for the Åland islands, which accounted for 14–19% of international passengers travel through Finnish ports in 2019–2021. Therefore, considering the data as a “sample” would be an understatement as the data represents most of the population.

Figure 3 presents turnovers of the port companies for 2019. It illustrates the dominant role of the Port of Helsinki, which is by far the largest port in terms of turnover.

The port size was categorized in a relative terms: the five largest ports in terms of turnover in 2019 were categorized as large, the next six as medium and the remaining seven as small. Business focus (passenger, freight or combined) categorization was categorized on the basis of transport volumes in 2019. The multipurpose vs. single-purpose attribute reflects whether the port was mainly serving one or few industrial entities (e.g. a steel or paper manufacturing plant) or whether its traffic was serving a mix of cargo and/or passenger flows. Some ports had imbalanced flows, with either export or imports dominating, while others maintained more balanced, two-directional flows. The freight type was also assessed as either bulk or unitized (cargo in containers and/or trailers). The bulk cargo included break-bulk, project deliveries and other non-unitized cargo. The data set was intentionally kept manageable by not distinguishing specific commodities as these tend to fluctuate in volumes based on market conditions. For example, grain prices affect not only grain exports and imports but also processed food products derived from grain. Similarly, the exports and imports of machinery may be affected by single industrial projects, such as the construction of a power station.

Port location is generally indicated by the company name (e.g. Port of Hanko Ltd. is located in the city of Hanko). Ports with turnovers above 12 MEUR in 2019 were categorized as large entities, while those ports with below 5 MEUR in turnover were categorized as small, and the rest of the companies were categorized as medium-sized ports. The main lines of business (freight, passenger or both) was assessed using data from the Finnish Transport Agency 2018. Most ports handled mainly freight, but a few had or a combination of freight and passenger traffic.

Ports that exclusively handled the transportation of goods for a single industrial entity were categorized as single-purpose ports, while those that served a wider customer base were categorized as multipurpose ports. Ports with less than one-third of their freight volumes in either export or import category were considered mainly import or mainly export ports, respectively. Based on annual reports from 2019 and Port Association data, ports were also classified by freight type: bulk, utilized or both. Bulk included project deliveries, break-bulk or other non-unitized cargo. Unitized cargo referred to containers and trailers, including trucks.

The annual financial statements were obtained upon request for research purposes from the Finnish Patent and Registration Office (PRH, 2022). The statements were aggregated and standardized into a uniform format (see Table 2). In essence, no adjustments were made to the statements and the figures reflect the accounting numbers as presented in the official statements. Since the accounting principles and financial reporting guidelines for Finnish limited companies are standardized, comparability is ensured. Only the data from 2019 to 2021 was used, as this period covers both the impact of the pandemic and the subsequent recovery.

The companies’ turnovers consisted of vessel call fees, waste management services and other port-related services such as mooring and unmooring vessels or the use of cranes. The fees vary from port to port and are usually tied to the vessel size and cargo type. Port companies also charge fees for trucks and trains entering the ports and rent land for cargo storage and warehousing.

For capital changes in balance sheets the ratio between equity and debt was measured in the anticipation that owners might have reacted to the crisis and declining turnovers by infusing more equity in port companies. Equity capital is usually cash-based by nature and hence a source of funds for the company. Debt capital can be raised by borrowing short or long-term.

Statistical testing was conducted using one-tailed t-tests to assess the differences between the means of different groups of ports. The t-test assumption indicates whether the means of two independent groups are significantly different from each other (Snedecor, 1946, pp. 76–88). The null hypothesis H₀ assumes that the group means are either equal or in the opposite direction of the presumption (μ₁ ≥ μ₂). Meanwhile, the alternative hypothesis H₁ assumes that μ₁ < μ₂ if H₀ is rejected, provided the risk of a Type I error is low (that is, rejecting H₀ when it is actually true). Multivariate analysis was also considered but rejected due to the small population and multiple groupings, resulting in ineffective analysis. The t-test statistic is calculated as follows:

where $x‾1$ and $x‾2$ are sample means, n₁ and n₂ the sample sizes, and

where $s12$ and $s22$ are sample variances. When t deviates sufficiently from zero, it indicates the probability of Type I error (p). The application of t-test assumes normally distributed variables. Normality was tested using the Shapiro–Wilk test at a significance level α = 0.1 for type I error. The Statistics Kingdom free online software package was used for technical analysis.

Business continuity profiles, following the logic shown in Figure 2, are summarized in Figure 4. The figure illustrates the average turnover decline in the turnovers of each identified port type. The arithmetic mean is justified since the business continuity analysis focused on decision-making units (DMUs), i.e. the port companies, rather than total turnover changes across the port sector. The decline, or drop, in the turnovers was measured as a turnover change between 2019–2020 and the recovery as the change between 2019 and 2021. The constant scale across the panels in Figure 4 allows for a visual comparison of the depth of the decline and strength of the recovery. Results demonstrated that most ports suffered turnover losses due to the pandemic. However, certain port characteristics indicated either resilience or vulnerability when considering the impacts on turnovers. Based on these descriptive statistics, some statistical tests were conducted in Section 4.4 (see also Table 4).

Most ports had not regained their pre-pandemic turnover levels of 2019 and remained below the zero line in 2021. However, a few exceptions were observed.

The first notable observation concerns the type of cargo the ports handled. Contrary to expectations, ports focusing primarily on unitized cargo (containers, trailers) suffered the most. Figure 5 shows that the three unitized-cargo ports lost, on average, almost a quarter (23%) of their turnovers in 2020 compared to 2019. By contrast, the nine-bulk focused ports lost only 5% on average. The difference is substantial. The remaining six ports handling both unitized and bulk cargo survived severe impacts, losing only 7% of their turnover on average between 2019 and 2020.

Smaller ports survived the pandemic better than their larger counterparts. The largest third of the group had a mean turnover change of less than −10% for 2019–2021, while small and medium-sized recorded only a minor decline of a few percentage points. For size and freight focus, the differences in means were statistically significant (see Section 4.5).

Further insights emerge when port types are clustered according to their turnovers losses between 2019 and 2020, and their recovery in 2021. The results are shown in Figure 6. Ports with import-focused traffic recovered from the pandemic particularly well. Some even managed to increase their turnovers during the shock year of 2020. In contrast, the unitized cargo ports, as noted above, experienced quite severe turnover decline and only modest, though positive, recovery. For instance, some of the unitized cargo ports lost roughly 30% of their turnover between 2019–2020 but recovered less than 10% between 2020–2021.

Ports with significant passenger volumes showed mixed outcomes. One port (marked by an arrow in Figure 6) lost almost 30% of its turnover during the decline and showed no visible recovery. This port primarily handled traffic to Estonia. Three other passenger ports performed much better. Two of them did not experience any decline and were even capable of increasing their turnovers both between 2019–2020 and 2020–2021. These ports were mainly handled traffic to and from Sweden, which maintained far fewer travel restrictions than Estonia. Moreover, Finland and other Nordic countries (Denmark, Norway and Iceland) were exempt from many restrictions (Government Offices of Sweden, 2020). Hence, keeping borders open also kept the transports flowing.

Thinking of the radical restrictions on passenger traffic, the impacts of the pandemic appeared mild on ports that also carried passengers. Only one passenger port (the Port of Helsinki) out of four experienced severe turnover impacts. Not surprisingly, the busiest passenger port recorded a notable drop in turnover. Passenger volumes declined from 12.8 million in 2019 to 4.7 million in 2020, a decrease of 63%. The impacts were not only market-driven but also the result of strict government policies to control the pandemic (Huoltovarmuusorganisaatio, 2022), which particularly affected the traffic between the Port of Helsinki in Finland and the Port of Tallinn in Estonia. This connection was (and still is) the most important link for the Port of Helsinki.

Another interesting observation arises when considering turnover drop and recovery simultaneously, as this can be regarded as one dimension of business continuity. Using the year 2019 as the benchmark, it is possible to draw the correlation between the turnover drop and recovery. The results are presented in Figure 7. There appears to be a relatively strong correlation (R² = 0.66) between the drop and recovery. Simply put, port companies that experienced a more severe turnover drop also recovered more poorly than others. While it may seem obvious that one would recover less quickly from a deep drop, the observation is, nevertheless, surprisingly clear. Another significant finding is that import oriented ports, shown in red color in Figure 7, proved the most resilient during the pandemic and were able to maintain their business continuity almost as if nothing happened.

The simple linear regression and statistics are as follows:

1. R = 0.813, R² = 0.661, indicating a very strong linear relationship between the variables; p-value is less than 0.001 for the entire equation and F-statistic is 31.23. Therefore, hypothesis H₁ is accepted (H₁ = there is a correlation)

2. Regression model: Ŷ = 4.0641 + 0.95X, where Ŷ is the modeled recovery level and X is the observed turnover drop. The slope coefficient (0.95) suggests an almost one-to-one relationship between drop and recovery; the p-value for the slope coefficient is < 0.001, and for the intercept term p = 0.147.

3. Due to the small sample size, the statistical power appears low (0.34): The power analysis, however, does not account for the fact that the sample represents a substantial share of the population (18 ports out of the 23), and covers about 80% of the transport volumes. Therefore, the low power is less critical as it would be in the case when sampled from larger or infinite population.

4. The Shapiro–Wilk test p-value equals 0.163 > 0.1, allowing us to assume that the data are normally distributed, though not with a strong margin of safety.

5. The data contains no outliers.

6. A free online statistical software (Statistics Kingdom) was used for the analysis.

One additional observation is worth noting. Although not statistically significant, there is anecdotal evidence suggesting that many larger ports were able to maintain business continuity less effectively than their smaller counterparts. This is shown from the plot in Figure 8 and further examined in Section 4.5.

One way to ensure business continuity in times of crisis is to financially strengthen the business entities. This typically achieved through capital infusions, creating buffers to withstand sudden losses of revenues. In the case of Finnish port companies, there was a clear tendency to react to the crisis by strengthening the financial solidity. Solidity was measured by the ratio between equity capital and debt capital (see Table 2). The solidity ratios for the ports are shown in Figure 9, illustrating changes from 2019 to 2020 and 2020–2021. Most port companies increased their solidity by injecting more equity capital into the company, but this is particularly evident among privately owned companies. The rationale for why private owners chose to take this approach was not directly observable from the documents.

There were quite a few companies (11) that continued to inject capital for two years in a row, while some companies (4) first increased equity and then decreased it or relied more heavily on debt financing, most likely because they saw that the pandemic was easing. Several companies opted for another strategy, using either their existing liquid equity capital or borrowing more to offset the decline in revenues. The two privately owned seaport companies (Inkoo and Tolkkinen) were those who most significantly increased their equity-to-debt ratios. One plausible explanation is that private owners were quicker and more responsive in ensuring business continuity, having faster access to internal capital sources.

The port management’s motivation to inject additional equity has not yet been examined, whether it was merely to secure short-term operational performance during the pandemic or to maintain the port’s reputation in the long term. Furthermore, whether BCPs or procedures has been implemented in the Finnish ports, and how mature they were in implementing them, has not been analyzed. It is likely that larger ports implemented more targeted BCM or BCPs for their most critical business activities or processes, while smaller ports may have relied on a single BCP (UNCTAD, 2023).

Other port attributes, such as specialization (single-purpose vs. multi-purpose) and primary line of business (passenger vs. freight), did not show any meaningful correlation with the business continuity that was the main focus of this measurement. The data were scattered without any identifiable patterns. The data are presented in Table 3, where the port companies are arranged in ascending order of business continuity (i.e. the weakest first and the strongest last). What can be stated with confidence is that the largest ports seemed to recover more slowly, while ports handling mainly bulk traffic indicated clearer business continuity.

There was one anomaly, however, that deviated from all other observed patterns: the increase in turnovers of private ports in 2020 compared with 2019 (see Figure 3). While there was a uniform decline in the turnovers across almost all ports and port types, the two private ports experienced an increase in turnovers in 2020 and then experience a decrease in 2021, while others were already recovering. This anomaly was largely driven by the bigger of the private ports (Inkoo), which may be related to its specialization in tramp shipping. Tramp shipping refers to irregular shipping using non-standard routes with no fixed schedules or port call itineraries. Typically, tramp vessels carry low-value bulk cargo that is not time-sensitive.

Tables 3 and 4 summarize the analysis results, including statistical tests that confirmed some of the descriptive statistics. Table 3 presents the descriptive data, where the ports are arranged in ascending order in terms of their business continuity. It was measured first by the decline in turnover between 2019–2020 and then by the recovery of turnovers between 2020–2021. The tests shown in Table 4 were run independently because the small sample size was divided into multiple nominal categories, which did not allow sufficiently powerful multivariate or rank-sum analyses. Some statistically significant observations were confirmed, however,

When the size categories (small, medium, large, measured by 2019 turnover) were reduced into two groups (smaller and larger), by splitting the data in the middle (N = 9 observations in each group), a one-sided t-test showed a statistically significant difference in the mean turnover change for 2019–2021. Recalling that the decline was measured between 2019 and 2020, and the recovery between 2020 and 2021. This result coincides with Figure 8, which illustrates how ports’ size appears to affect financial resilience, at least in the regard of the crisis. The significant p-value for this test was p = 0.014, indicating that only 1.4% risk of error in accepting the alternative hypothesis (H₁ = smaller ports recovered better).

The second statistically significant difference in resilience was explained by freight type, as shown descriptively in Figure 5. Testing showed significance when bulk ports were compared with ports handling either mainly unitized cargo or both types. A one-sided t-test was also applied to test whether the mean turnover change from 2019 to 2021 was higher for passenger and freight ports than for bulk-only ports. The p-value for this test was p = 0.049.

Testing the difference between multipurpose and single-purpose ports (the latter linked to a major industrial production facilities), produced a borderline result. The hypothesis that single-purpose ports recovered better than multi-purpose ports were just barely rejected with a p-value of 0.104. This is such a close case that it would be unreasonable to mechanically conclude that the hypothesis should not be accepted.

The Finnish seaport companies recovered reasonably well from the pandemic. The year 2020 saw the most severe impact, by the year 2021, however, most port companies were already on the path to recovery. The European Maritime Safety Agency (EMSA) reported in its 2021 COVID-19 impact report (European Maritime Safety Agency, 2021), that Finland was among the countries most affected by the pandemic. However, when examining the actual impacts on financial performance and business continuity, it can be concluded that, fortunately, the worst scenarios were not realized. Some port companies survived the shock surprisingly well, and for instance ports focused on import-oriented traffic were less affected. This was also observed in Leviäkangas et al. (2023), although their study only examined turnover decline, not recovery. The reason why export-oriented ports and those bi-directional ports did not cope so well compared to import-oriented ports still requires further investigation. Additionally, smaller ports were found to perform better in terms of business continuity and recovery. Therefore, the intuitive assumption that larger port entities and those with greater business diversification would have stronger business continuity, is not supported by the observations of Finnish port companies. The two privately owned seaport companies were quickest to react to the pandemic, as evidenced by capital structure changes between 2019 and 2020. They increased their equity capital relative to debt capital more than others but were also quick to reverse or lower the equity infusion a year after. The likely explanation is that the pandemic was presumed to be better under control, and the number of uncertainties decreased as experience was gained.

The statistical evidence was partly quite clear. When testing the mean turnover change for 2019–2021 (first plunging between 2019–2020 and then bouncing back 2020–2021) several results were significant or close to significant. The ports that recovered best were smaller ports, single-purpose ports and ports handling only bulk freight. This suggests that specialization may be a strength in port operations, while diversification and size may not. Another alternative explanation is that bulk handling in industrial ports is just too critical to be restricted. Industrial ports exist to serve industry, and the owners of these facilities have significant influence on port operations. Without the industrial facility or facilities, the port might not even exist. At the same time, it is obvious that the reliance on single or some industrial facilities creates a different kind of vulnerability. Individual ports also affected the statistical results. For example, Port of Helsinki – being the largest port and the one that lost the most – had a substantial impact on the overall analysis. However, excluding Port of Helsinki from the dataset would not have yielded more valid results, especially since 18 out of 23 seaports were included. In other words, this study covers almost the entire population, thus outliers should not have been an issue in the first place.

Regarding the limitations of the results, other factors that may have contributed to the decline in traffic and, consequently, turnovers. For example, war in Ukraine or the obstruction of Suez Canal, both of which disrupted global traffic. However, compared to the obvious impacts of the pandemic, these other factors can be considered marginal. The war in Ukraine began in 2014 and is still ongoing, while the Suez Canal obstruction occurred in March 2021, when recovery from the pandemic had already started. However, it is important to acknowledge that isolating of the impacts attributed solely to the pandemic is somewhat impossible. The key point is whether the pandemic’s impacts can reasonably be assumed to have been the decisive factor, thereby enabling measurements.

The Finnish “port system” has often been criticized as fragmented, comprising too many of small port entities. However, there may also be arguments in favor of having many small ports. For instance, for each municipality and industrial facility, the ports they own are of strategic importance, and for municipalities, they are also important source of revenue (Rönty et al., 2011). Tapaninen (2015) analyzed port architecture in Finland and concluded that capacities to handle different types of cargo and commodities overlapped, with four or five ports able to handle similar type of cargo. This clearly provides flexibility in exceptional situations. If such research would extend further, Tapaninen (2015) could offer an alternative way of categorizing different types of ports and contribute additional information. It should be underlined that the observations made in this study do not constitute as “conclusive evidence”. Conclusive evidence would likely require cross-country studies and the identification of more detailed explanatory variables behind the observations. Moreover, it is evident that changing the attributes used to classify ports might yield entirely different explanations for successful (or unsuccessful) business continuity. Therefore, more research is required, while acknowledging that most generalizable truths begin with observations. On the other hand, the evidence provided by the observations, measurements and tests carried out in this study should not be downplayed either. The observations are quite clear, there is a substantial sample (almost an entire population) and some statistical evidence is significant.

When considering the implications for various stakeholders and business actors, including port owners, shipping lines, port operators, stevedoring and forwarding providers, as well as workers, the results yield only indirect implications. The two most obvious stakeholders can be identified. First, owners of larger seaport companies should carefully consider whether size provides protection in a crisis such as pandemics or whether risks could be somehow diversified or mitigated. This would require an entirely new analysis that delves deeper into the operations of the seaports. Since such an investigation is practically difficult, given that companies protect such information as confidential business matters – there is little chance that meaningful, comparative findings could be achieved. In turn, shipping lines may reconsider which ports they prefer to visit and with which they wish to establish strategic collaborations. Intuitively, one might assume that shipping lines prefer larger ports with wider repertoire of services, however, in light of this study, there may be grounds for alternative thinking. Even so, without further analyses, these remain prospective future research questions.

While port companies are just one of business actors in the maritime transport business ecosystem, they are also an excellent indicator of the transport economy and industrial activity, since their revenues depend directly on the volume of vessel calls. Therefore, port business continuity reflects economic resilience in a broader sense, as it is directly linked with the continuity of the trade flow continuity. Although seaports management models have been studied to some extent, the results of this study did not address any particular management model or tool for dealing with unexpected events such as the pandemic. Rather, it focused on identifying resiliency patterns among different types of seaports. Nevertheless, business continuity management, alongside with risk management and quality management, was identified by Loh and Thai (2014, 2015, 2016), shortly before the pandemic, as prospective management tool for dealing with unexpected disruptions.

The theoretical implications of this study are indirect, since no explicit theory was either tested or proposed. However, a prospective theoretical framework that aligns with this analysis is the Panarchy Theory. Panarchy theory suggests that siloed and disciplined-specific approaches are insufficient to address the complexities faced by modern socio-ecological systems (Sundstrom et al., 2023). Traditional approaches can easily lead to sub-optimal decisions and unexpected side effects, since the causalities in complex networks are largely unknown and dynamic. The same reasoning applies to techno-economic systems in general. Pursuing economies of scale inevitably creates risk and vulnerability concentrations, as larger scale typically brings in larger networks and more points of connection, thereby increasing the likelihood of disruptions. Less interconnected systems and less complex networks, by contrast, are easier to manage and likely reduce the probability of having disruptive domino effects.

The theory distinguishes between two types of systems: those that are small, agile, fast, flexible and uncomplicated and those that are large, rigid, slow, established and interconnected systems. It suggests that capabilities of reacting and adapting are stronger in the former type. The findings from the Finnish seaports support this line of thinking: smaller ports and specialized single-purpose ports proved more resilient. Many of these ports continued their operation without severe disruption from pandemic, whereas larger, highly connected and diversified ports faced greater challenges.

A valid question remains whether formalized and standardized business continuity management supports flexibility and agility or whether it creates managerial and organizational rigidity. The idea of BCM is somewhat deterministic: in anticipation of sudden adversity, formal processes and guidelines are designed to respond in a predefined manner. Yet the presence of BCM does not necessarily equate to resilience, which penetrates deeper in the qualitative and quantitative characteristics of organizations. Perhaps inconclusively, yet observably, the Finnish ports’ experience of surviving and recovering from the pandemic seems to suggest the opposite. At the very least, the results hint formalization and pre-defined procedures do not always work – the world might just be too complex and full of surprises.

PL conceptualized the study, developed the methodology, collected the data and drafted the manuscript. Background research was done by BW and ZZ. The final manuscript was written by all authors, with review and editing done by ZZ. All authors have read and agreed to the published version of the manuscript.

All data used in this research are freely available, upon specific request, for research purposes from the Finnish Patent and Registration Office PRH. Clinical trial number: not applicable.

MSc Pekka Räty from The Uusimaa Centre of Regional Development and MSc Shahid Hussain from The University of Oulu are acknowledged for their consultation regarding statistical analyses.