The Coast Guard must maximize its effectiveness given its limited resources. Marine inspector and current Coast Guard Commandant Admiral Linda Fagan recently stated that the “Coast Guard is experiencing a workforce shortage. We are struggling to recruit the people we need to hire into our ranks.” Although this issue is not isolated to any single mission, the marine inspection workforce has recently been the object of a Government Accountability Office (GAO) report.1
The GAO Report
The maritime sector contributes approximately $5.4 trillion annually to the U.S. economy. The Coast Guard’s marine inspection workforce ensures commercial ships are safe to operate. After a series of boiler explosions resulted in numerous casualties, this responsibility was initially given to the Justice Department in 1838 and later to the U.S. Steamboat Inspection Service in 1871. In 1942, the Steamboat Inspection Service and its responsibility were transferred to the Coast Guard.2 Today, per the GAO, the marine inspection workforce comprises more than 725 men and women. Their responsibilities include ensuring U.S.- and foreign-flagged vessels meet statutory safety requirements and that maintenance and repairs are completed to minimize risk to the environment, waterways, and people.
The GAO report, which was released in 2022, found:
The supply of marine inspectors has consistently not met the estimated need. . . . The Coast Guard has not established performance measures with targets for its marine inspection workforce improvement plan and associated initiatives that would help identify desired outcomes and provide a means to measure how its efforts help close workforce gaps over time.
The GAO found that the service has a total shortfall of more than 400 marine inspectors, a gap of approximately 40 percent in estimated workforce needs.3 The report’s main conclusions reinforce the need to fix the workforce shortage and incorporate performance markers to improve mission effectiveness. However, the workforce is also facing challenges mitigating risk on vessels that use complicated technologies to maximize efficiency and reduce emissions.
Establishing Measures of Effectiveness
To improve mission effectiveness, today’s marine inspection workforce must focus on its founding mission: protecting people, property, and the environment. However, reliance on unsupported beliefs and biases should be avoided to put the workforce in a position to succeed. Good data must be a primary consideration for decision making at all levels.
The foundation of a data-driven framework is built on an adequate measure of effectiveness (MOE). For years, the marine inspection program has lacked a useful MOE. Marine inspectors prevent disasters at sea by identifying and documenting instances of noncompliance with applicable laws and regulations. Although issuing deficiencies to vessels for noncompliance has always been a core workforce function, it has not been incorporated adequately as a measure of inspection effectiveness.
Other Coast Guard missions have more easily adopted useful MOEs. It is common to see the Coast Guard publicize the number of lives saved or people assisted during search-and-rescue missions. In marine safety, effectiveness is often measured by the number of inspections conducted. However, this is like measuring the effectiveness of a cutter by counting underway hours.
The marine inspection community does count “detentions” as a measure of effectiveness. A detention occurs when a vessel is so grossly out of compliance that the marine inspectors prevent it from operating until the deficiencies are corrected. This is good, but it must be expanded to include all instances of noncompliance. Using detentions as the sole measure of effectiveness results in two categories of vessels: Those that have been detained, and those that have not. Such overly broad categories are not sufficient for an accurate evaluation of risk, particularly in an industry that relies so heavily on precise operations. An effective use of data should result in many different risk categories or rankings.
Predicting NonCompliance Rates and Locations
Establishing the MOE would be an important first step in implementing a risk-based framework, because it could guide the analysis of data. The marine inspection program has been collecting data on inspections and deficiencies for more than two decades. Deficiency data is broken down by ship system (e.g., lifesaving), component (e.g., life raft), and condition, along with other properties. With an established MOE, data could be analyzed to separate valuable characteristics from those that are less valuable. With an established MOE, a regression analysis—a tool widely used in industry—can be used to evaluate the data to separate valuable characteristics from less important ones to find out which characteristics have the biggest effect on an output.
For example, if a business would like to predict next month’s sales numbers, the sales numbers would be the output, which could be affected by any number of different inputs. These inputs could range from sales promotions to other quantifiable external factors that may have an effect on sales. Regression analysis sorts these factors to recognize which inputs play a big part in affecting the output. This process often reveals that a factor not previously considered or less heavily weighted did, in fact, have a strong effect on the output. Conducting an analysis in this manner allows for an accurate prediction of future outputs. Harvard Business Review described this as the “go-to method in analytics.”4
Many factors can affect deficiencies in a vessel examination, which is the primary output for marine inspections. These factors include, but are not limited to, operator, owner, vessel age, length, and tonnage. Deficiencies per exam should align with the MOE. By using a discrete output such as deficiencies per exam, along with the information collected in the marine inspection database, this process can be replicated to find deficiencies per exam broken down by ship system. Although this additional step is not difficult, its importance would be critical, as it would reveal not only which ships are at an increased risk of noncompliance, but also which ship systems are likely to be found noncompliant. This information would allow the workforce to optimize resources.
The goal of this process is to put the Coast Guard’s workforce in the best position to ensure the safe operation of the Marine Transportation System (MTS). Once the mathematical analysis is complete and the workforce has a better understanding of which ships in their ports carry a higher risk of noncompliance, resources can be better allocated to the mission. At field units, inspectors can be shifted daily based on seniority, experience, expertise, and training considerations to best ensure compliance.
Because of operational demands, field units can be limited in how they allocate resources on any given day. The frequency and scope of inspections are often set at the program level with consideration given to statutory requirements. The incorporation of data, however, could play a pivotal role in setting policy on the scope and frequency of inspections. In 2021, the Coast Guard altered the frequency of small passenger vessel inspections based on predicted risk.5 This was a significant step toward achieving a more data-driven workforce. There is room to build on this initiative though, as passenger vessels accounted for only 21.8 percent of the overall fleet subject to Coast Guard inspection.6
Using regression analysis, vessels not predicted to have high rates of noncompliance could be inspected no more than required by law, and their inspections could be limited in scope. This approach would reduce staff hours and increase effectiveness across the workforce. The scope of inspections could be continually analyzed, iterated, and clarified to ensure optimal use of resources.
Certain inspections can be delegated as needed to qualified third-party organizations and classification societies. For example, in 2022, 16.6 percent of exams of foreign-flagged gas carriers found deficiencies, while 35.06 percent of exams of foreign-flagged vessels carrying refrigerated cargo identified deficiences.7 Statutory authority mandates that the Coast Guard inspect gas carriers on an annual basis. However, if this authority were delegated to a classification society, the Coast Guard could more frequently inspect other vessel categories with higher rates of noncompliance (i.e., refrigerated cargo carriers). There is precedent for this, as the Alternate Compliance Program and Towing Safety Management System, among other initiatives, delegate specific statutory authorities and functions to approved organizations.
Incorporating data as outlined above would provide the additional benefit of promoting consistency across the inspection workforce. Based on the number of predicted deficiencies per inspection, individual unit trends and tendencies would be more easily identifiable. For example, if 85 percent of Coast Guard units conducting marine inspections fell within plus-or-minus 10 percent of their predicted deficiencies per inspection figure, management and oversight could be shifted to the 15 percent of units that fell outside these performance markers. Corresponding audits that measure critical nonconformities would result in overall program improvement. When significant nonconformities are not found, the regression analysis could be updated to consider the new information. The iterative nature of this process makes it sustainable.
Long-Term Resource Allocation
The Coast Guard’s sector-staffing model is discussed heavily in the GAO report and is used to assign billets to units based on the number of inspections they conduct. This results in a gross man-hour figure that is divided by the number of man-hours a Coast Guardsman can provide, ultimately yielding a billet allocation. This model assigns billets based on where inspections are conducted, not necessarily where high rates of noncompliance are found.
To allocate resources to address risk, the model should consider a mix of deficiencies per exam and inspections conducted because of statutory requirements. This can be done by assigning additional man-hours for each deficiency issued. Alternatively, the program can adopt additional inspection types to account for those done on vessels with high rates of noncompliance. Over time, adopting this holistic, data-driven approach would allow for the reallocation of resources to best meet areas of continued non-compliance.
Operationalizing data in this fashion has shifted from something that is “nice to have” to an operational necessity for protecting the MTS. Private companies have analyzed marine inspection statistics to identify “key port-specific aspects.”8 This information allows companies to see which ports are more prone to issue deficiencies and to which ship systems deficiencies are most often issued. Well-intentioned companies and operators can use this information to help their ships improve compliance with applicable safety regulations. However, it would be naïve to believe that other companies are not using this same information to avoid ports with higher historical rates of issuing deficiencies, or to make temporary fixes to more serious problems to pass inspections.
The technology used in a data-driven framework for marine inspections is available, proven, and understandable. However, changes to resource allocation and a substantial reduction in risk to the MTS requires a much wider adoption of data-driven practices. A data-driven marine inspection workforce would allow for a deeper understanding of commercial fleets, as well as a more effective and empowered workforce.
1. Government Accountability Office, Enhancements Needed to Strengthen Marine Inspection Workforce Planning Efforts, GAO-22-104465 (Washington, DC: GAO, 12 January 2022).
2. Barbara Voulgaris, “From Steamboat Inspection Service to U.S. Coast Guard: Marine Safety in the United States from 1838–1946,” media.defense.gov.
3. Government Accountability Office, Enhancements Needed.
4. Amy Gallo, “A Refresher on Regression Analysis,” Harvard Business Review, 4 November 2015.
5. U.S. Coast Guard Office of Commercial Vessel Compliance, “Small Passenger Vessel Risk-Based Inspection Program,” 14 June 2021.
6. U.S. Department of Homeland Security, U.S. Coast Guard, Flag State Control in the United States: 2021 Domestic Annual Report (Washington, DC).
7. U.S. Department of Homeland Security, U.S. Coast Guard Office of Commercial Vessel Compliance, Port State Control in the United States: 2022 Annual Report (Washington, DC).
8. “Port State Control (PSC) Services,” www.dnv.com.