NOAA Technical Memorandum NMFS NE 211
Trends in Selected Northeast Region Marine Industries
by Eric M. Thunberg
NOAA National Marine Fisheries Service, 166 Water Street, Woods Hole, MA 02543
Print
publication date July 2008 ;
web version posted April 21,2009
Citation: Thunberg EM. 2008. Trends in Selected Northeast Region Marine Industries. NOAA Technical Memmorandum NMFS NE 211; 107 p.
Download complete PDF/print version
ABSTRACT
The Northeast Fisheries Science Center provides scientific assessments and management advice for managed fishery resources in the “Northeast region” ranging from the US/Canada border in Maine to Cape Hatteras, NC. This study examines the performance and relative contribution of marine industries in coastal states bordering this Northeast region over time. These states include the New England states of Maine, New Hampshire, Massachusetts, Rhode Island, and Connecticut. The “mid-Atlantic region” is defined for study purposes as New York, New Jersey, Delaware, Maryland, Virginia, and North Carolina. For purposes of this study, the Northeast region marine-based economy identifies 14 industrial sectors classified under the North American Industry Classification System and/or the Standard Industrial Classification system, depending on year. In addition to a combined marine-based sector, trends in performance of these sectors are reported separately as well as in three industry groupings: seafood commerce (commercial fishing, seafood dealers, seafood processors, and retail fish markets), recreational boating (boat building, boat dealers, marinas, and sightseeing excursion services), and water transportation (shipping, ship building, marine cargo handling, port operations, and navigational services). Measures of performance include total employment, number of establishments, the location quotient, and the Herfindahl index. The location quotient provides a relative measure of regional specialization while the Herfindahl index measures changes in industry concentration.
The principal source of data used in this study was annual County Business Patterns data for the years 1986-2005. Methods to estimate employment where data have been suppressed because of confidentiality concerns are described. For commercial fishing, nonemployer statistics from the US Census Bureau and quarterly employment data from the Bureau of Labor Statistics were used.
The marine sector represents about 0.5% of total employment in the Northeast region. Employment growth has been positive, increasing at about the same rate as employment in the general economy. However, important differences exist in performance among the three marine subgroups and across states. Employment has been declining in both the seafood commerce and the water transportation subgroups, albeit at an average annual rate of less than 1%. Overall employment growth has been positive only because employment growth in the marine recreational boating subgroup has more than offset the employment declines in the seafood commerce and water transportation sectors. This also implies that employment in the marine recreational boating subgroup has been expanding at a rate greater than that of the general economy. Among the Northeast states marine sector employment has been declining in Maryland, Maine, New York, and Virginia. In each of these states positive growth in marine recreational boating employment has not been sufficient to offset declines in other marine subgroups.
INTRODUCTION
Although the principal focus of social and economic analysis of fisheries regulation is on fishing and fishing-related entities, the marine economy consists of a broader range of activities that take place in the coastal zone. Pontecorvo et al. (1980) defined the marine economy as including any activity that utilizes the ocean in a productive process or that exists because the demand for the sector’s output is due to some attribute of the ocean. On the production side, Pontecorvo et al. distinguished between activities that extracted living (fish) or nonliving (minerals) resources from those that directly utilized the ocean (shipping). On the demand side, a distinction was drawn between activities that are attributable to the ocean (seafood processing) and activities that were located in proximity to the ocean (a gift shop). This conceptual framework has been used to estimate the economic activity supported by Large Marine Ecosystems (Hoagland et al. 2005) and to examine the role of the marine economy in the state of Massachusetts (Georgianna 2000; Donahue Institute 2006).
As conceived by Pontecorvo et al. (1980), proper delineation of the ocean sector would require a change in data collection protocol to what was then the Standard Industrial Classification (SIC) system to apportion activity in specific sectors between ocean and nonocean components[1]. For example, ice is used both by commercial fishing vessels and, among other things, by consumers to chill beverages. Apportionment to the marine economy would require an accounting of how much ice was used in its various marine and nonmarine uses. Such a change was never adopted. This means that the data upon which most studies of marine economies depend contain a mixture of marine and nonmarine related establishments (i.e., businesses). Depending on which sectors are included, the economic activity attributed to the marine environment may be under- or overestimated. For purposes of this study, only sectors that (1) are associated with extractive uses of the marine environment, (2) are directly associated with extractive uses, or (3) make direct use of the ocean were included.
Changes in the marine economy may be attributable to many different factors. For fishing and fishing-related industries, change may be driven by resource conditions and the regulatory actions taken to rebuild a fishery. These changes may also be driven by demand for alternative uses of shore-front property. The objective of this report is to document trends in numbers of establishments and level of employment for selected marine industries. In each case indicators of performance are developed to identify sectors that may be increasing or decreasing or those sectors whose structure or relative role in a local or regional economy may be changing over time. Spatial differences in performance among states as well as coastal counties within states are identified.
The remainder of this report contains four major sections. The first section identifies the marine sectors that were selected. The second section describes the data used in the report. Special considerations made because of changes in industrial classification as well as data suppression are discussed. The third section identifies the performance indicators, their calculation, and their interpretation. The last section identifies specific groupings of marine sectors to evaluate how commercial fishing and fishing-related sectors as a whole may be changing as compared to recreation-based or shipping/water transportation sectors. Finally, results for individual marine sectors are also presented.
MARINE SECTORS
The study objective is to examine how the relative contribution of fishing and fishing-related industries to the Northeast regional economy has been changing over time. In this study the “Northeast region” is defined as the New England states of Maine, New Hampshire, Massachusetts, Rhode Island, and Connecticut, and the “mid-Atlantic region” states of New York, New Jersey, Delaware, Maryland, Virginia, and North Carolina. These New England and mid-Atlantic states border the fishery resources assessed by the Northeast Fisheries Science Center.
In this study the range of marine industries considered is somewhat narrower in scope than the more general treatments of the contribution of the marine-related sectors as a whole that were done by Georgianna (2000) and the Donahue Institute (2006). The latter study grouped 56 different industrial sectors based on North American Industy Classification System (NAICS) codes into five major sectors: commercial seafood, marine transportation, coastal tourism and recreation, marine science and technology, and coastal construction[2].
Neither the SIC nor NAICS-based industrial classification systems provides the level of detail necessary to distinguish between many types of establishments that primarily service fishing or fishing-related activities from establishments that serve nonfishing clients (e.g., welding, trucking, ice manufacturing). For this reason, construction of a marine sector for purposes of this study was limited to a smaller set of industries that are more likely to be predominately associated with direct use of coastal waters or directly related to those uses. These industries include seafood harvest; seafood processing; seafood wholesale; retail fish markets; boat building; boat dealers; marinas; marine sightseeing; ship building; water transportation of passengers and cargo; and water transportation services, including marine cargo handling, dock and pier operations, and navigation services to shipping. These sectors were classified into three groupings: commercial seafood, marine recreational boating, and water transportation (see Table 1). Note that these groupings bear some similarities to the groupings used in the Donahue Institute report (2006).
DATA
Data for this study come from County Business Patterns (CBP) from the US Census Bureau for calendar years 1986-2005 (US Census Bureau 2008). County Business Patterns is an annual series that provides detailed employment and payroll data by industrial classification for counties and states. These data are obtained through several sources including annual surveys conducted by the Census Bureau and the administrative records of the Internal Revenue Service and the Social Security Administration (US Census Bureau 2007). The unit of observation is an establishment, which is defined as a single physical location or place of business. In cases where multiple activities are carried out under the same ownership, all activities are classified under a single establishment. The industrial classification for that multiactivity establishment is based on its major activity. This means that the reported number of establishments may underestimate the total number of establishments that may be engaged in a particular kind of activity. For example, seafood businesses may process fish or shellfish and may also act as wholesale distributors or buyers/sellers of unprocessed seafood. Any such establishment would be assigned to a single industrial classification (either processing or wholesale trade) depending on which activity was the larger source of revenue. Note that this also means that from one year to the next, the industrial classification of an establishment may change as the major activity changes.
Covered establishments include any business with paid employees during the mid-March pay period. That is, the CBP data represent a point-in-time estimate rather than an annual estimate of employment throughout the year. For businesses with a strong seasonal pattern of employment, the CBP data would tend to underestimate total employment unless March is an average or above average season. Additionally, sole proprietorships or self-employed individuals are not covered nor are railroad, domestic service, agricultural production, government, or foreign workers.
Data collected over the time series were based on two different major industrial classification systems: 1986-1997 were SIC-based while 1998-2005 data were based on the newer NAICS. Additionally, the time series covers two sets of revised industry codes for both SIC and NAICS. Specifically, 1986-1987 data were based on the 1972 SIC codes while 1988-1997 data were based on 1987 SIC codes, and 1998-2002 data were based on 1998 NAICS codes while 2003-2005 data were based on the 2002 NAICS codes. In some cases no changes were made to the types of activities that were included in an industrial classification while in others substantial revisions were made. For sectors where the SIC and NAICS codes are consistent, all available years were used (see Table 2). Where the changeover from SIC to NAICS classification systems created a discontinuity in the time series, only data from 1998-2005 were used.
In accordance with federal statute, employment and payroll data may be suppressed where disclosure would violate confidentiality restrictions. However, neither the total number of establishments nor their distribution by employment size class is considered disclosure.
The amount of suppressed data differed by industrial sector and level of aggregation (i.e., state or county). Data suppression was particularly prevalent for industrial sectors such as mining which have only a small presence in the Northeast region. Similarly, data suppression was more prevalent at the county level than at the state level.
Proration of suppressed data followed a procedure similar to that described in Minnesota IMPLAN Group (2000). Both SIC and NAICS industrial classification systems are organized at major industry levels as well as by subgroupings within each major industry where employment data at each level of aggregation sum to the totals at the next higher level of aggregation. For a SIC-based classification scheme, this means that employment for all 4-digit codes add up to the 3-digit codes, the 3-digit codes add up to the 2-digit codes, and so on. An estimate of employment is obtained by multiplying an adjustment factor Ai to an initial estimate of employment Êi where Ai is calculated as:
[1] Ai = [Ei+1 – ΣEi]/ΣÊi,
where i denotes the level of aggregation for the suppressed sector, i+1 denotes employment at the next higher level of aggregation, E denotes unsuppressed employment and Êi denotes the initial estimate of employment for suppressed data. The initial estimate for Êi was obtained by taking advantage of the fact that numbers of establishments are reported by employee size class (1-4, 5-9, 10-19, 20-49, 50-99, 100-249, 250-499, 500-999, 1,000-1,499, 1,500-2,499, 2,500-4,999, 5,000+). For all but the terminal size class, employment in each size class was estimated by multiplying the midpoint of the size class interval by the number of establishments. Since the upper bound for the terminal size class is indeterminate, a different approach was required.
To derive an initial estimate for the 5,000+ size class, one approach would be to make an assumption about the average number of employees then multiply by the number of establishments to get an estimate of total employment in the terminal size class. Unfortunately the choice of average employees is arbitrary and could introduce a downward or upward bias to the adjustment factor if employment is over- or underestimated. For suppressed employment data, a suppression flag is included where each flag corresponds to a range estimate of total employment. These ranges and suppression flags include 0-19 (A), 20-99 (B), 100-249 (C), 250-499 (E), 500-999 (F), 1,000-2,499 (G), 2,500-4,999 (H), 5,000-9,999 (I), 10,000-24,999 (J), 25,000-49,999 (K), 50,000-99,999 (L), and 100,000 or More (M). Of these intervals only the last five (I-M) were used to estimate employment in the 5,000+ size class since any other suppression flags would correspond with total employment of less than 5,000. Procedurally, employment in the 5,000+ size class was estimated by subtracting the sum of the estimated employment in all size classes below 5000+ (described above) from the midpoint of the total employment interval. For example, the midpoint estimate for the suppression flag K would be 37,499. Since actual employment is uncertain, this estimate may still under- or overestimate the adjustment factor, but rather than being set at an arbitrary average, the initial estimate of employment in the terminal size class is allowed to vary by industry and is grounded in reported total employment.
The initial estimate of total employment (Êi) is obtained by summing the estimated employment in all size classes. Note that it would have been possible, and simpler, to obtain Êi by using the midpoint of the employment range corresponding to the data suppression flag. This method was not selected because it does not take into account potential changes in employment as the number of establishments change from one year to the next. That is, total employment may fall within the same range but may increase/decrease as establishments enter/exit. The method used for this study permits these changes to be reflected in the employment estimate.
The initial employment estimate is adjusted as in [1] from the bottom up to obtain an adjusted estimate of total employment. For example, ship and boat building (NAICS 33661) is a 4-digit NAICS industry that is further broken out into two 5-digit subsectors (ship building and boat building). To illustrate the procedure, consider that New Jersey employment at the 4-digit NAICS code for ship and boat building was 2,030 people in 2005. However, the separate 5-digit codes for boat building and ship building were suppressed. The reported number of establishments in each employee size class interval for each sector is shown in Table 3. Multiplying the number of establishments in each size class by the size-class midpoint and then summing these counts results in an initial employment estimate of 825.5 and 1,376 for boat building and ship building respectively.
Applying [1] to these data yields an adjustment factor of 1.054 (2320/825.5 + 1376). Note that in this example, 2,320 corresponds to the unsuppressed Ei+1 in [1], (825.5 + 1376) corresponds to ΣÊi, and ΣEi is set to zero since there were no unsuppressed sectors at the 5-digit level. Multiplying the adjustment factor by the initial employment estimate produces a final estimate of 870 employees in the boat building sector and 1,450 employees in ship building. If, for example, employment for the 4-digit code was also suppressed, then employment for NAICS 33661 would have to be estimated by proration to total employment at the 3-digit level before adjusting the suppressed 5-digit data. Once all suppressed data have been adjusted from bottom to top, a second pass is made from top to bottom to make any final adjustment to ensure that all data add to the overall total.
Because a large amount of data had been suppressed for fishing establishments in the CBP data, employment for fishing was estimated by using Quarterly Census of Employment and Wages (QCEW) data from the Bureau of Labor Statistics (BLS), and from nonemployer statistics from the Census Bureau. The QCEW data are based on the quarterly census of employment and wages, and this data collection program is administered jointly by the US Department of Labor and state Employment Security agencies. As a quarterly census, the QCEW data represent annual employment, unlike the CBP data which are point-in-time estimates. Data reported in the QCEW program include employment and wages covered under state unemployment insurance programs and federal workers covered by the Unemployment Compensation for Federal Employees program. Like the CBP data, reported information is subject to confidentiality restrictions, but since establishments by size class are not reported, it is difficult to estimate employment for nondisclosures. Data for the entire time series from 1986-2005 were obtained from the BLS, where all data from 1986-2000 were based on the SIC classification system and from 2001-2005 were based on NAICS codes[3]. No adjustments to the data were necessary since there is direct correspondence between the two industrial classification systems for commercial fishing.
The QCEW data includes individuals who are covered by unemployment insurance, whereas the fishing industry includes a large number of individuals who may be self-employed crew or owner-operators. These sole proprietorships are covered under the nonemployer series available from the Census Bureau. The nonemployer series contains numbers of individuals and total receipts for any business that file federal income tax but that did not report any paid employees. Data are reported for calendar years 1997-2005 on a NAICS basis. Total fishing employment was then the sum of QCEW employment and the number of sole proprietorships.
PERFORMANCE INDICATORS
Four performance indicators were used: employment, number of establishments, location quotient, and Herfindahl index. Employment and number of establishments provide a direct indicator of the presence or absence of a marine sector in a given region as well as how either may be changing over time. The location quotient provides a measure of the relative role of a marine sector within a given region, while the Herfindahl index provides a measure of the firm size structure for a given industry.
The location quotient (L) is calculated as:
[2] Lij = [Eij/EiR]/[Ej/ER],
where Eij is employment in sector i in region j, EiR is total employment in sector i across all regions, Ej is total employment in region j for all sectors, and ER is total employment across all regions and sectors. The numerator in the location quotient is equivalent to the employment share in region j for sector i, while the denominator is essentially the average employment share for region j across all sectors. If the employment share for sector i exceeds its regional average, the location quotient will be greater than 1 and industry i may be said to play a proportionally larger role in the regional economy relative to other regions. From the standpoint of impact assessment, the relative impact in a region with a location quotient above 1 is likely to be larger compared to regions where employment in an affected marine sector may be proportionally lower.
The Herfindahl index (H) for a given sector is calculated as:
[3] H = Σ(ei)2/E2,
where ei is employment in establishment i and E is total employment in the sector. The Herfindahl index is bounded between 1 and 1/N. That is, the index takes on a value of 1 if there is a single establishment and 1/N if every establishment is of equal size. The Herfindahl index provides a measure of industry concentration. In general terms, a high Herfindahl index is indicative of an industry that is dominated by a few large firms whereas a low index is indicative of an industry with many small firms. A Herfindahl index that is increasing over time suggests that industry consolidation may be occurring whereas a decline in the index suggests that the industry is becoming more dispersed. Note that the Herfindahl index calculated herein is only a rough approximation of industry consolidation for two reasons. First, the CBP definition of an establishment does not take into account the potential for ownership control over multiple establishments at different sites. This means that the Herfindahl index calculated for this report cannot be used to measure changes in consolidation in ownership control or market power, the more traditional use of the index. Second, a precise measure of the Herfindahl index requires employment data for each individual establishment, which are not available in the CBP data. Instead, an approximation of the index was derived by assuming that the number of employees in each establishment in a given employment size class was equal to the size class midpoint. Thus this approximation of the Herfindahl index (Ĥ) was calculated as:
[4] Ĥ = Σ(êi)2ni/[Σêini]2,
where êi is the midpoint estimate for the employee size class i and ni is the number of establishments in the size class.
As noted above, the value of the Herfindahl index depends on the number of establishments and the size of each establishment. Kelly (1981) showed that the Herfindahl could be decomposed into “fewness” and “standardized dispersion” effects. These effects provide a means for determining whether a change in the Herfindahl is predominately due to a change in the number of establishments or to a change in the size structure of the sector. Specifically, the fewness effect is measured by 1/N while the standardized dispersion effect is measured by the ratio of the observed variance in establishment size and the variance if the sector were a “virtual monopoly,” that is, the sector is dominated by a single large establishment. The dispersion effect is bounded between 0 and 1. As the dispersion effect approaches 1, the sector is trending toward virtual monopoly. Conversely, the sector is trending toward uniform size as the dispersion effect approaches zero.
MARINE INDUSTRY TRENDS
Although data for some marine sectors were available for the entire time series, a consistent time series for all sectors was only available from 1998-2005. These data include the combined seafood commerce, marine recreational boating, and water transportation industry groups defined herein. Note that because of data suppression in the QCEW series during calendar years 1998, 1999, and 2000 for seafood harvest in Maine and Massachusetts in particular, total marine industry establishments and employment were underestimated in these two states. This means that Northeast region totals for those years were also underestimated but that the relative magnitude of the underestimate was less pronounced. For purposes of reporting, the number of establishments only includes reporting units that had paid employees. Sole proprietorships included in the nonemployer series for seafood harvesting were included in employment totals.
The number of combined establishments in the seafood commerce, marine recreational boating, and marine transportation sectors totaled an average of about 6,700 from 1998-2005 (Table 4). These establishments represented less than 0.5% of total establishments in all industrial sectors in the Northeast region (Table 5). Across states, New York had the largest number of marine industry establishments and averaged more than 1,700 establishments from 1998-2005 although the marine sector only accounted for 0.35% of total establishments in the state. By contrast, marine sector establishments represented nearly 2% of establishments in Maine, highest among Northeast region states.
The share of total marine industry establishments was at least 25% in New York from 1998-2005 (Table 6). In Massachusetts the share of establishments was at least 18% in 1998 and from 2001-2003, but declined in 2004 and in 2005 to 17.1%. For the most part, however, the share of marine industry establishments has been quite stable with only a small increase in Maine and small decreases in Massachusetts and New Jersey.
Combined marine industry mid-March employment averaged 137,000 employees in the Northeast region from 1998-2005 (Table 7). Total employment was a time series low of 133,700 in 2002 but increased in each of the last three years reaching 140,000 employees in 2005. Virginia had the largest number of employees (an average of 32,000 employees) primarily because of the state’s dominance in the ship building sector.
Total marine industry employment was approximately 0.5% of total Northeast region employment ranging from 0.52% in 2001 to 0.57% in 1998 (Table 8). Because of the large ship building sector in Virginia, total marine industry employment was at least 1% of total employment in all sectors of the Virginia economy. Marine sector employment was also at least 1% of total state employment in Rhode Island, and in Maine marine employment represented about 3.6% of total state employment.
The share of Northeast region marine sector employment was largest in Virginia averaging 23.7% from 1998-2000 (Table 9). Maine’s share of regionwide marine sector employment was as high as 15.3% in 1999 but has been on a slight declining trend to 13.5% in 2005. By contrast, the employment share in Connecticut has been increasing in every year since 2002. In other states the regionwide employment share has fluctuated without any notable trend from one year to the next.
As a measure that takes into account both differences in sector employment and total employment shares among parts of a larger region, the location quotient provides a means for comparing the relative role of specified economic sectors across states. For example, New York’s marine employment share was about three times that of Rhode Island’s. However, since the share of total employment in Rhode Island in 2005 was much less than that of New York, the Rhode Island marine sector location quotient was 2.0 while New York’s was 0.4 (Table 10). Note that the location quotient provides only a comparative measure of relative importance among states; it does not provide a measure of the magnitude or importance of any given sector to a state. The location quotient exceeded one in Connecticut, Maine, Rhode Island, and Virginia. This means that marine sector employment is overrepresented in these states relative to their total employment share.
The marine sector includes three subgroups representing commerce in seafood products: marine recreational boating activities and sectors related to shipping and ship building. Seafood commerce represented similar proportions of both numbers of establishments and employment in the Northeast region from 1998-2005 (Table 11). That is, the share of establishments and employment in the seafood commerce marine sector subgroup was approximately one-third of total marine sector establishments and employment. Note that the share of seafood commerce establishments has been declining from 35.1% in 2002 to 30.2% in 2005 while the share of employment declined from 34.4% in 2002 to 32.7% in 2003 but held steady in both 2004 and 2005.
Marine recreational boating establishments accounted for over one-half of all marine sector establishments in the Northeast but accounted for between 17-22% of employment. The opposite was the case for the water transportation marine subgroup. Primarily because of the underlying structure of the ship building sector within the water transportation subgroup, the subgroup accounted for about 12% of total establishments but represented more than 45% of total employment.
In the following sections, trends in establishments and employment for each of the marine sector subgroups are reported. In each case, trends for the subgroup as a whole as well as trends for each sector that comprise the subgroup are discussed.
Seafood Commerce
The seafood commerce subgroup includes commercial fishing, seafood dealers, seafood processors, and retail seafood markets. The available time series for each of these sectors differed, and trends for each are reported for the longest period of time possible. However, trends for the seafood commerce sector subgroup were available for all sectors only for the period 1998-2005. As noted previously, available data for the commercial fishing sector include establishments with hired employees and sole proprietorships. For reporting purposes, the former were included as establishments to be consistent with the CBP definition of establishment while the latter were included with total seafood commerce employment.
The total number of seafood commerce establishments in the Northeast region ranged from a high of 2,409 in 2002 to a low of 2,001 in calendar year 2005 (Table 12). The decline in number of establishments was most notable from 2002-2003 and from 2004-2005. In each of these cases, the number of seafood commerce establishments fell by 163 and 229 reporting units, respectively. Note that this pattern was evident across all states, so the decline in seafood commerce establishments was a regionwide phenomenon not associated with a precipitous decline in any one state. At least some of these declines were attributable to data suppression in the QCEW data used to estimate commercial fishing establishments and employment. However, other than in Maine and Massachusetts, the number of establishments subject to suppression was quite low, so any underestimation in numbers of establishments at the Northeast region level from 2001 onward would be small.
Total Northeast region employment in the seafood commerce subgroup was 47,341 employees in 1998 and 45,817 employees in 2005 (Table 13). Note that the effects of underestimation of employment in the seafood harvesting sector are less than that of establishments because in states other than Massachusetts fishing employment associated with sole proprietorships was 80% and in most cases more than 90% of commercial fishing employment. This means that even though some data suppression occurs in one or more states from 1998-2003, underestimation of total employment in the Northeast region and of employment in each state for that matter is likely to be a concern only in years (1999 and 2000) where Massachusetts data were suppressed.
In a manner similar to that of total establishments, total employment declined from 2002-2003 and from 2004-2005. However, the proportional reduction in employment was substantially less than that of the change in establishments, suggesting potential employment loss from the reduced number of establishments was at least partially absorbed by those that remained.
Since data suppression is likely to misrepresent seafood commerce employment in Massachusetts, employment shares were calculated for calendar years 2001-2005 (Table 14). Employment share has been declining in Virginia from 10.5% in 2001 to 8.8% in 2005 while employment share has been increasing in Massachusetts from 18.6% in 2002 to 22.5% in 2005. Employment shares in other states have been relatively stable although Maine’s share of seafood commerce employment has been gradually eroding since 2002.
In the Northeast region, more than half of seafood commerce employment is composed of commercial fishing employees (Table 15). Compared to 2001, commercial fishing employment in 2005 declined by only one percentage point while seafood dealer employment declined by 4.6 percentage points. Offsetting decline in fishing and seafood dealer employment shares was an increase in the employment share of both the seafood processing and the retail seafood market sectors. Among some states the pattern in changes in employment shares matched that of the region as a whole (Massachusetts, for example), while others did not. In Maine, the commercial fishing employment share increased in 2005 compared to 2001, but the seafood processing employment share declined. The following provides a discussion of trends in each of the seafood commerce sectors.
Commercial fishing NAICS 1141
The commercial fishing sector is composed of establishments engaged in the harvest of finfish, shellfish, and other marine products taken from their natural habitat. Any sector engaged in husbandry activities is included in the animal aquaculture sector (NAICS 1125). Animal aquaculture was excluded from this report because it is classified with agricultural production which is not covered by CBP. The CBP data were not adequate to characterize fishing employment for two reasons. First, a large portion of available data were suppressed since many engaged in the fishing industry are considered self-employed. Second, commercial fishing has a strong seasonal component meaning that a substantial amount of employment would be missed during the mid-March survey period used by CBP. For these reasons fishing employment was estimated by using QCEW and nonemployer statistics. Both of these data sources provide an estimate of total annual employment. The former captures establishments subject to payroll taxes while the latter captures self-employed or sole proprietorships. Data for both series were available from 1997-2005.
Massachusetts had the largest number of reporting units (Table 16) and the most employees (Table 17) in all years where data were not suppressed. These data indicate that the number of Massachusetts reporting units was larger in 2005 than in 1998 but has remained relatively constant from 2003-2005. However, the number of fishing employees has declined recently from 1,701 in 2003 to 1,682 in 2005, a reduction of about 1%. By contrast the number of reporting units and fishing employees in Maine has increased in consecutive years from 2001-2005.
In addition to number of reporting units and employees, the QCEW data report total annual wages paid to employees. Adjusting for inflationary effects by using the Consumer Price Index (CPI) indicates that average annual wages paid to fishing employees have been increasing in most states, particularly from 2003-2005 (Table 18). For example, average annual wages in Maine increased from $21,640 in 2003 to $23,825 in 2005. Similarly, average annual wages paid to fishing employees in Massachusetts has increased to a time series high of $38,333 in 2005. Compared to state per capita personal income, adjusted by the CPI, average fishing wages were above the statewide average in Maine, Massachusetts, and Rhode Island (US Bureau of Economic Analysis 2009). By contrast, average fishing wages tended to be 40-60% lower than statewide per capita personal income in Delaware and North Carolina and 10-20% less than statewide amounts elsewhere.
The number of sole proprietorships reporting fishing income averaged 22,300 from 1997-2005 (Table 19), representing almost 88% more people engaged in fishing as a sole proprietor either as captain, crew, or sole operator of a fishing boat compared to wage-based employment. Over time, the total number of sole proprietorships engaged in fishing has been declining at an average rate of 0.6% per year. The average annual change in sole proprietorships was positive in Delaware, New Hampshire, New York, and New Jersey while annual change in sole proprietorships was negative in all other states. This negative change was at least 1% per year in the states of Maryland, North Carolina, and Virginia.
The number of fishing sole proprietorships in Maine was nearly twice as great as that of any other state, accounting for 27-29% of all such proprietorships from 1997-2005 (Table 20). Massachusetts also accounted for at least 20% of fishing sole proprietorships while the remaining New England states combined accounted for less than 10%.
Gross receipts for Northeast region fishing sole proprietorships adjusted by the CPI were nearly constant at about $22,000 (Table 21). Annual receipts were above the Northeast region average in Maine, Massachusetts, New Hampshire, New Jersey, and Rhode Island. Compared to average QCEW wages paid to fishing employees, gross receipts to sole proprietorships tended to be slightly higher in Maine and New York, and substantially higher in New Jersey, North Carolina, and Virginia. By contrast, wage-based income tended to be higher than sole proprietorship receipts in Massachusetts and Rhode Island.
An estimated location quotient was obtained by adding QCEW and sole proprietorship data. Note that this estimate would be biased downward in cases where QCEW data were suppressed. This downward bias is likely to be larger where QCEW data were suppressed since wage-based employment was somewhat higher (Maine and Massachusetts, for example) compared to others.
The location quotient was consistently greater than one in the states of Maine, Massachusetts, Maryland, North Carolina, and Virginia (Table 22). These five states accounted for approximately 70% of total Northeast region fishing employment. A location quotient above one means that the commercial fishing employment share in a given state was higher than the state’s total employment share, relative to other states in the Northeast region. Thus a reduction in commercial fishing employment in these states would be expected to have a proportionally greater impact on the state economy compared to other states.
The location quotient for states was calculated relative to the Northeast region as a whole. However, fishing employment may be comparatively more important to coastal areas within states than to the state as a whole. To examine the relative importance of fishing in coastal areas, the location quotient was recalculated for coastal counties in each state. The selected coastal counties correspond to the coastal regions developed by Steinback and Thunberg (2006). These coastal regions include a total of 143 counties and independent cities (see Figures 1-11). The location quotient for each county was calculated by using the fishing employment and total employment in the state as a base. For example, the location quotient for Washington County, ME, was the county’s share of fishing employment in all coastal counties in Maine, divided by Washington County’s share of total employment in all Maine coastal counties.
The location quotient exceeded one in 63 of the 143 coastal counties (Table 23). On average, these counties accounted for 75% of total fishing employment in the Northeast Region. As noted previously, counties where the location quotient exceeds one provide an indication of where changes in fishing employment would be expected to have a larger impact on the local economy as compared to counties where the location quotient was less than one. Note that counties where the location quotient exceeds one include those with more than 1,000 people working as sole proprietorships or wage employees (Hancock, Knox, and Washington Counties in Maine and Bristol and Essex Counties in Massachusetts) as well as counties where there were fewer than 50 people were engaged in fishing (Middlesex County, CT, Richmond County, NY, Middlesex and Salem Counties, NJ, and Wicomico County, MD). These large differences in employment levels highlight the difference between measuring the role of fishing in coastal economies in terms of absolute numbers as compared to a relative measure like the location quotient. That is, even seemingly low levels of fishing employment may play a large role in coastal counties that have a small population and a comparatively smaller labor force.
Seafood wholesale NAICS 42446
The seafood wholesale sector includes establishments that purchase fish and seafood directly from fishing vessels as well as establishments engaged in the distribution of fish and seafood that may have been processed. These establishments do not include wholesalers that distribute either canned or frozen packaged seafood.
Employment in the seafood wholesale sector was increasing at an average rate of 1.1% per year from 1986-1996 (Table 24). Regionwide employment began falling in 1997 and has declined in almost every year. That is, total wholesale dealer employment has been declining at an average annual rate of 3.3% from 11,683 employees in 1996 to 8,532 in 2005. Over the same period employment attrition in the wholesale dealer sector declined at a higher average annual rate in Delaware, Maine, Rhode Island, and Virginia while New Hampshire was the only state with positive average annual employment growth from 1997-2005.
The number of wholesale seafood establishments increased from 1,003 in 1986 to 1,324 in 1996, a time series high (Table 25). Since 1996 the number of establishments declined to fewer than 1,000 businesses in both 2004 and 2005. The same general trend has been occurring across all Northeast region states although the average annual decline in number of wholesale seafood distributors from 1997-2005 was higher than the regionwide average in all New England states except Maine. By contrast, attrition in numbers of wholesale seafood distributors has occurred at lower average annual rates in most mid-Atlantic states.
The location quotient exceeded one in the states of Massachusetts, Maine, and Rhode Island in every year from 1986-2005 (Table 26). In relative terms, a change in seafood dealer employment in these New England states may be expected to have a larger impact than in other Northeast region states. The location quotient was greater than one in Virginia up to 1995 but has since fallen below one, indicating a decline in the share of wholesale employment in Virginia relative to the state’s share of total regional employment.
The Herfindahl index depends on the number of establishments as well as the size distribution of the industry. Since the seafood dealer sector is predominantly composed of establishments with fewer than 100 employees, the value of the index as well as changes in the index are largely determined by the number of establishments. This is why the Herfindahl index for the seafood dealer sector in the Northeast region as a whole is substantially lower than the index value for individual states (Table 27). This is also why the index value increases so much in Delaware, as the number of seafood dealers has declined to only two establishments in 2005. Since 1997 the average annual change in the Herfindahl index has been positive in all states except Maine. In most cases the change in the Herfindahl has been due primarily to changes in the size structure of the seafood dealer sector as the dispersion effect exceeded the fewness effect at least for states other than Delaware.
Of the 143 coastal counties from Maine to North Carolina, the location quotient was greater than one in 55. Seafood dealer employment (Table 28, see Table 29 for numbers of establishments) in these counties ranges from fewer than 10 to several hundred, highlighting the fact that the relative importance of marine industries within a region or state is not solely measured by numbers of employees alone. For example, Cumberland County, ME, has the largest number of seafood dealer employees in the state but also accounts for the largest share of total employment in Maine such that the location quotient for the county is less than one. Note that this does not mean that seafood dealer employment is unimportant to Cumberland County, it only means that in relative terms, other Maine counties have a higher dependence on seafood dealer employment.
Seafood processing NAICS 3117
Seafood processing is a subgroup included in the industry group of food manufacturing. Under the NAICS seafood processing is further broken down into seafood canning (NAICS 311711) and fresh and frozen seafood processing (NAICS 311712). The former includes canning, smoking, curing of seafood as well as factory ships that harvest and can seafood on-board. The latter includes fresh and frozen processing of finfish and shellfish as well as processing of marine fats and oils. From 1986-1997 these activities were included in SIC codes 2091 (canning), 2092 (fresh, frozen processing), and 2077 (animal and marine fats and oils). The two classification systems made no change to the classification of establishments engaged in seafood canning, but establishments engaged in processing marine fats and oils formerly included in SIC 2077 were reclassified to NAICS 311712 creating a discontinuity between SIC 2092 and 311712 at least at the national level. As will be noted below, no evidence of such a discontinuity between SIC and NAICS years was apparent in the Northeast region.
For purpose of analysis, seafood canning and fresh/frozen processing were combined into a single higher subgroup (NAICS 3117 and SIC 2090). Since this also resulted in higher numbers of total establishments in each state, the number of instances of data suppression that required estimation of employment was reduced. In 1986 there were 309 seafood processing establishments (Table 30). The number of establishments declined by an average of 2.7% from 1986-1996 after which the number of seafood processors stabilized at about 235 establishments through calendar year 2000. In 2001 the number of processors in the Northeast region declined to 228 and fell again to 216 establishments in 2002. Although the number of processing establishments rebounded to 224 in 2003, the number of processors declined in both 2004 and 2005 to a time series low of 212 establishments in 2005.
The general decline in seafood processing establishments resulted in a decline in seafood processing employment in the Northeast region by an average annual rate of 5% from 1986-1996 (Table 31). By 1995 this declining trend appears to have been halted as processing employment hovered around 8,000 people from 1996-2001. Processing employment fell to a time series low of 7,227 employees in 2002. However, even as the number of seafood establishments declined in both 2004 and 2005, seafood employment actually increased in these years to 8,611 in 2005.
Performance of seafood processing employment and number of establishments varied considerably across states. Connecticut, New Jersey, and New Hampshire were the only states that showed an increasing trend in both processing establishments and employment both from 1986-1997 and from 1998-2005. By contrast, Maine was the only state that experienced a decline in employment and in the number of processing establishments over the same time periods. Since 1996 the average annual change in number of processing establishments was negative in the states of Maryland, North Carolina, and Virginia. However, the average annual change in total seafood processing employment was actually positive in these states, suggesting that at least some of the processing employment was being absorbed by other processors as establishments were closed.
The location quotient for seafood processing exceeded one in all years for Delaware, Massachusetts, Maryland, Maine, and Virginia (Table 32). In New Hampshire and Rhode Island, the location quotient was above one in most years and has remained above one in both states in every year since 1996. Notably, the location quotient in Maine has been declining because of its falling share of regional processing employment. However, Maine’s processing employment share is still much larger than Maine’s share of total regional employment. This means that in relative terms, a change affecting processing employment would have a larger impact in Maine than in other states.
The seafood processing Herfindahl index for the Northeast Region remained quite low in all years, a finding consistent with a sector with a large number of establishments not dominated by a few large establishments (Table 33). Over time the regional Herfindahl index has increased, averaging an annual change of 5.2% and 6.0% respectively from 1986-1996 and from 1997-2005. The majority of this change was accounted for by changes in dispersion rather than by changes in the number of establishments. That is, standardized dispersion has been increasing, indicating a slight trend toward larger establishment size. Note that standardized dispersion and the Herfindahl index itself are still very low, so the processing sector is far from becoming dominated by a few large entities, at least at the level of individual establishments.
Because of very low numbers of establishments in Connecticut and Delaware, the Herfindahl index was 1.0 or nearly so for much of the time series. The Herfindahl index for New Hampshire declined from 0.78 in 1986 to 0.23 in 1994 because of increasing numbers of establishments as well as a trend toward a more uniform distribution of establishment size. More recently the New Hampshire processing Herfindahl has increased, but only slightly. Over the time series of analysis, the processing sector in North Carolina exhibited the largest change in the Herfindahl index.
In 1986 North Carolina had 32 seafood processing establishments with a Herfindahl of 0.05. At least through 2001 the North Carolina Herfindahl index increased modestly to 0.08, which is still characteristic of a processing sector comprised of similarly sized establishments. Since 2001 the Herfindahl index has risen dramatically from 0.12 in 2002, to 0.21 in 2003, 0.36 in 2004, and to 0.41 in 2005. Since the number of establishments over these years has not changed very much, the majority of this increase has been attributed to a trend toward greater dominance by a small number of larger establishments. For example, in 2002 the largest processing establishment had fewer than 100 employees. In 2004 and 2005 the largest processing establishment employed 250-499 employees. Although the Herfindahl index has been increasing in most other states, the shift in the size structure of individual processing establishments that has occurred in North Carolina does not appear to be occurring elsewhere.
Among coastal counties in the Northeast region, there were 43 where the location quotient was greater than one in most, if not all years from 1998-2005. These counties accounted for the majority of processing establishments and employment in their respective states, and for 70% of the total number of establishments (Table 34), and 81% of employment (Table 35) in the Northeast region.
Seafood retail NAICS 44522
Establishments engaged in selling processed or cured finfish and shellfish to retail customers compose one of three sectors included in the specialty food stores industry group (NAICS 4452). In addition to retail fish markets, this group includes meat markets and fruit and vegetable markets. Under the SIC classification system, seafood markets and meat shops were combined, making it impossible to track trends in seafood retail establishments and employment prior to 1998.
The number of seafood retail establishments has increased in the Northeast region from 809 in 1998 to 1,035 in 2005 (Table 36), an average annual change of 3.7%. The number of retail seafood establishments in New York was at least twice that of any other state, ranging from 302 establishments in 1998 to 392 seafood markets in 2005. In all other states, the number of seafood retail markets was 100 or more in only Massachusetts and New Jersey. The average annual change in seafood markets was positive in all states and was highest (8.6%) in Maine, growing from 28 establishments in 1998 to about 50 over the past 3 years.
Employment in the seafood retail market sector changed by an average of 5.8% from 3,148 in 1998 to 4,620 employees in 2005 (Table 37). The average annual change in seafood market employment was positive in all states ranging from 3.8% in North Carolina to almost 12% in Connecticut. New York accounted for almost one-third of regionwide employment while the combined employment in New York, New Jersey, and Massachusetts accounted for approximately 60% of total retail seafood employment.
The location quotient for seafood retail markets exceeded one in all years in Delaware, Maryland, Maine, New York, and Rhode Island (Table 38). Note that New York’s location quotient was only slightly above one even though it had twice the number of establishments and employment of any other state because New York’s share of total regionwide employment was only slightly less than its retail seafood market employment share. Conversely, Maine’s location quotient was at least 2.0 because Maine’s retail seafood employment share was twice that of its share of total regionwide employment. Thus, a change affecting retail seafood markets would be expected to have a comparatively larger impact in Maine than it would in New York, but the impact on New York would still be expected to be larger than that of states where the location quotient was less than one.
The Herfindahl index for the Northeast region retail seafood market sector was very low; this measure is indicative of a sector characterized by a large number of establishments not dominated by a small number of large entities (Table 39). The regionwide Herfindahl index has increased but by only a very small amount. Among states, the average annual change in the Herfindahl was positive in Massachusetts, North Carolina, New Jersey, New York, and Virginia. However, in these states the value of the Herfindahl index itself as well as the standardized measure of dispersion is suggestive of an industrial structure of predominantly small sized establishments at least in terms of employment.
There were 47 coastal counties that had a location quotient for the retail seafood market sector that exceeded one in most, if not all, years. Note that these counties were determined by comparing coastal counties within a state. This accounts for the large range in terms of establishments and employment reported in Table 40 and Table 41, respectively.
Marine Recreational Boating
The marine recreational boating subgroup is composed of establishments engaged in building boats for personal use, retail boat dealers, and marinas. Additionally, the subgroup includes establishments that provide water-based excursions, which includes activities such as whale watching, recreational fishing, and harbor or sightseeing cruises. The number of Northeast region establishments engaged in these activities has increased from 3,407 in 1998 to 3,696 establishments in 2005 (Table 42). The number of establishments has increased in each year since 1999 by an average of 1.8% per year. Among states New York had nearly two-thirds more marine recreational boating establishments compared to New Jersey, the state with the second-most establishments. However, the number of boating-related establishments in New York changed at a lower average annual rate (0.8%) than was the case for Delaware (4.9%), Rhode Island (3.7%), and Maine (2.0%).
The marine recreational boating subgroup employed almost 24,000 people in the Northeast region in 1998 (Table 43). Employment in the subgroup has been trending upward at an average annual rate of 3.3% to 30,000 people in 2004 and 2005. Among states, North Carolina experienced the largest growth in recreational boating subsector employment at an average of 8.4% per year. Although the average annual change in employment was positive in all states, employment levels in six of the ten states declined in 2005 compared to 2004 levels.
In excess of two-thirds of the marine recreational boating subgroup employees were employed in a mid-Atlantic state (Table 44). Within mid-Atlantic states, employment has been shifting to North Carolina as the state’s employment share has increased from 14.6% in 1998 to 20.6% in 2005.
Regionwide employment in boat dealers and marinas averaged one-third each of the marine recreational boating subsector employment (Table 45). The balance was composed of 8% employment in the marine excursion sector and 27% in the boat building sector. Employment shares of these marine sectors varied across states. For example, boat building was nearly 50% or more of the marine recreational boating subgroup employment in Maine, North Carolina, and Rhode Island.
Boat building and repair NAICS 336612
The boat building sector comprises establishments engaged in building watercraft that are not built in shipyards and that are primarily intended for personal use. Boats may include dinghies, motorboats, rowboats, and sailboats. Boat building was also included as a separate sector under the SIC system but was subdivided into two sectors under the NAICS (boat building and other personal household goods; NAICS 81149), a division which created a break in the time series. The number of boat building establishments has been increasing over time. These establishments increased from 283 in 1998 to 316 in 2005 (Table 46) at an average annual change of 1.7%. The increase in establishments was lower than the relative increase in employment (Table 47). This difference in growth rate between employment and number of establishments is reflected in an increase in the number of employees per establishment from 21.6 in 1998 to 27.6 in 2005.
Total employment in the boat building sector has been increasing since 1998, growing from 6,104 in 1998 to 8,715 in 2005; this is an average annual increase of 5.3% (Table 47). Much of this growth was accounted for by increased boat building employment in North Carolina which more than doubled from 1,655 in 1998 to 3,917 in 2005. Boat building employment in North Carolina, Maryland, Maine, New Jersey, and Rhode Island combined accounted for over 80% of the Northeast region total in 1998 and increased to over 90% of the total in 2005.
As reflected by a location quotient greater than one, the share of boat building employment in Maryland, Maine, North Carolina, and Rhode Island was proportionally greater than each state’s respective share of total employment in the region (Table 48). That is, boat building employment in these four states plays a proportionally larger role in total state employment than in other Northeast region states. Therefore, a change in boat building employment would have a proportionally greater impact in Maryland, Maine, North Carolina, and Rhode Island than in other states.
The Herfindahl index for the boat building industry was quite low (0.02 or 0.03 in all years), suggesting an industry that has few large establishments (Table 49). Note that the Herfindahl index was higher in some states than in others. However, since the lower bound of the index is affected by the number of establishments, comparisons across states are not valid unless each state has approximately the same number of establishments. For example, Massachusetts and Maryland had nearly identical number of establishments from 2003-2005. Over these years, the Maryland index was larger than the Massachusetts index, suggesting that the boat building industry in Maryland tended to be composed of larger establishments over these years than those in Massachusetts. Similarly, the number of establishments in New Jersey, New York, and Rhode Island was similar in all years, yet the New Jersey Herfindahl index was higher, indicating an industry that is composed of larger establishments than those in either New York or Rhode Island. The Herfindal index was nearly constant in most states from 1998 through 2005. However, the index has generally declined in Maryland and increased in Virginia, suggesting a trend toward small firm size in Maryland while the industry has tended toward consolidation in Virginia.
Businesses engaged in boat building may be more concentrated within certain counties and less so in others. Given the large number of coastal counties throughout the Northeast region, the location quotient was used to identify a subset of counties where boat building may be a proportionally larger share of local employment compared to other coastal counties within a state. Note that the location quotient can exceed one even in cases where employment in the sector of interest is very small. This circumstance occurs in counties with a small labor force relative to the total employment in the region.
Boat building employment and number of establishments varies considerably in Northeast region coastal counties. Consistent with the generally higher level of boat building employment at the state level, employment in coastal counties in Maine, Maryland, North Carolina, and Rhode Island was generally larger than that of coastal counties in other states (Table 50). In Maine, the boat building sector was concentrated in the downeast counties of Hancock and Washington County as well in the midcoast counties of Lincoln and Knox. The number of establishments in these Maine counties, Hancock County in particular, also tended to be higher than elsewhere in Maine and other coastal counties throughout the Northeast region (Table 51).
The boat building industry in Maryland is clustered around the Chesapeake Bay with Anne Arundel County located on the western side of the bay while all other establishments were clustered in a nearly contiguous string on the eastern side of the bay including counties from Kent to Wicomico Counties. Note that of these Maryland counties, the majority of boat building employment was in Wicomico County even though the number of boat building establishments was higher elsewhere, in Anne Arundel and Kent Counties in particular. That is, the boat building industry in Wicomico has a small number of large establishments while other Maryland counties tend to have a larger number of smaller establishments.
Boat building employment in North Carolina is geographically dispersed from north to south with Chowan, and Dare Counties in the north; Beaufort, Craven, and Carteret Counties on the central coast; and Brunswick County to the south. Pitt County also accounted for a substantial share of coastal county boat building employment. However, the boat building location quotient in Pitt County did not exceed one, indicating that relative to other coastal counties boat building employment was a lower proportion of the Pitt County work force. In 2005 boat building employment in Craven County was about 25% of the coastal county total but was around 10% in each of the other four counties. In terms of establishment size, the boat building industry in Carteret and Dare Counties had at least twice as many establishments indicating that boat building in these counties consists of smaller establishments as compared to Chowan, Beaufort, Brunswick, and Craven Counties.
In Rhode Island the boat building industry was clustered in the adjacent counties of Bristol and Newport. These two counties accounted for nearly all of the boat building employment in the state. Since 1998, boat building employment has been shifting somewhat from Newport to Bristol County as the share of employment in Bristol County has been increasing relative to Newport County. The number of establishments has also been increasing in Bristol County but has been declining in Newport County.
Boat dealers NAICS 441222
The boat dealer sector includes establishments that retail new and used boats. Dealers may also offer repair services as well as retail outboard motors, boat trailers, accessories, and other marine supplies. The number of boat dealer establishments increased from 1,040 in 1986 to 1,334 in 2005 (Table 52). Growth in the boat dealer sector averaged 0.9% per year from 1986-1996 and grew at nearly 2% per year from 1997-2005. Among Northeast region states, the average annual change in number of establishments was negative from 1986-1997 in Connecticut, Massachusetts, and Rhode Island. However, from 1997-2005 the number of boat dealer establishments increased in all states, ranging from 6.7% per year in Rhode Island to less than 1% in New Hampshire, New York, and Virginia.
Employment in the boat dealer sector increased each year from 1986-1988 before entering a period of decline from 1989-1993 (Table 53). Since 1994, employment in the boat dealer sector has been trending upward, averaging an increase of 3.6% per year from 1997-2005. Virginia was the only state with a negative average annual change in boat dealer employment over the same time period. However, this period includes the change in Virginia employment from 1996-1997 which was -36%. Omission of this one year results in an average annual change of 4.4%. Note that in most states the average annual change in employment from 1997-2005 exceeds the average increase in boat dealer establishments, indicating that employment growth has resulted from a combination of increasing numbers of dealers and adding employees at existing establishments.
The location quotient for boat dealers was consistently greater than one in Delaware, Maryland, Maine, New Hampshire, and Rhode Island (Table 54). In relative terms, the boat dealer sector plays a proportionally larger role in these states compared to other Northeast region states. For this reason a change in the boat dealer sector may be expected to have a larger proportional effect in these states even though total boat dealer employment was larger in absolute terms in states such as New York and North Carolina.
The Herfindahl index for the Northeast region as a whole, as well as across states, was very low from 1986-2005 (Table 55). Additionally, the index has changed relatively little over time although the average annual change in the index indicates some decline in the Herfindahl because of increasing numbers of establishments. That is, the underlying size structure of the industry has remained relatively stable even as new establishments have been added.
The location quotient was greater than one for 59 of the 143 coastal counties included in the analysis. These counties accounted for 53% of the Northeast regional total number of establishments (Table 56) and 57% of regionwide employment (Table 57). Since the location quotient was calculated for coastal counties on a state-by-state basis, these counties include coastal counties with a large number of establishments and employees as well as counties with only a few.
Marinas NAICS 71393
Marinas include establishments engaged in operating dockage or mooring facilities for owners of pleasure craft. These establishments may also provide related services such as retailing fuel and boating supplies and providing maintenance, repair, and rental of pleasure boats. The number of establishments offering marina services increased from 984 in 1988 to 1,511 in 2005 (Table 58). On average the annual change in number of marinas was 5.8% from 1988-1996 and was slightly negative (-0.1%) from 1997-2005. Much of the growth in numbers of marinas occurred from 1988-1996. Since 1996, the number of establishments has ranged, without trend, between 1,400 and 1,500 establishments. This general pattern is evident across states as growth in the number of marinas was positive in all states from 1988-1996 while the average annual change in establishments was lower and in some cases (Maryland, North Carolina, and New Jersey) was slightly negative.
Regionwide marina employment has increased from 6,487 in 1998 to 9,004 in 2005 (Table 59). Growth in marina employment trended upward throughout the time period even as the number of establishments was relatively stable from 1996-2005. This suggests that, on average, the number of employees per establishment has been increasing. Among states expansion in marina employment was highest in New York and Virginia as employment in the latter nearly doubled from 532 in 1988 to more than 1,000 employees in 2005. By contrast marina employment in Delaware expanded from 1996 through 2002 but has dropped to less than 100 in 2003, 2004, and 2005.
The marina sector location quotient exceeded one in Connecticut, Maryland, Maine, and Rhode Island (Table 60). In New Hampshire the location quotient exceeded one in most years although barely so.
The marina sector Herfindahl index was very low for the Northeast region as a whole and for all states except for Delaware (Table 61). From 1988-1996 the average annual change in the Herfindahl index was negative in the Northeast region. This declining trend was due to the expansion in number of establishments that was occurring over that time. The average annual change from 1997-2005 was positive because of some small changes in the size structure of Northeast region establishments. However, these changes had little impact on the Herfindahl index as it remained constant from 2003-2005 at 0.002.
The Herfindahl index in Delaware was higher than any other state because of a combination of a small number of marina establishments and a change in the size structure of the industry which was particularly notable from 1996-2002. Prior to 1996, mid-March employment did not exceed 19 for any marina establishment in Delaware. From 1996-2002 there was at least one establishment that had anywhere from 50-500 employees. This means that a single establishment accounted for large proportion of total Delaware marina employment in these years, hence the relatively high Herfindahl index.
The location quotient exceeded one in 53 of the 143 coastal counties included in the analysis. These counties included counties with a large number of establishments (Table 62) and employees (Table 63) as well as counties with only a few of either. Among coastal counties, the number of establishments exceeded 50 in only Suffolk County, NY, Anne Arundel County, MD, and Ocean County, NJ. These three counties alone accounted for about 15% of total marina employment in the Northeast region.
Sightseeing water transportation NAICS 487210
Establishments included in the scenic and sightseeing water transportation sector provide local excursion services that typically return to the point of origin on the same day of departure. These services include such activities as party or charter fishing, whale watches, dinner cruises, and harbor or lighthouse sightseeing tours. In the Northeast region the number of establishments offering these sightseeing opportunities increased in every year from 2000-2005 (Table 64). On average, the number of establishments increased 2.8% per year from 1998-2005. Among states, the number of water-based excursion establishments was highest in New York and New Jersey although the average annual change in number of establishments in these two states was less than 1%. By contrast, the number of establishments more than doubled in 2005 compared to 1998 in Delaware and nearly doubled in Rhode Island. In other states, the change in number of establishments was more modest, ranging from 2.3% per year in Virginia to 5.8% in New Hampshire.
Although the number of water-based excursion establishments increased, the number of employees changed little from 1998-2005 (Table 65). Growth in employment averaged less than 1% per year. Among states average annual growth was negative in Connecticut, Massachusetts, and New Jersey. Elsewhere the average annual change in employment was positive although employment in 2005 was below 1998 levels in Maine and North Carolina.
The location quotient exceeded one in every year from 1998-2005 in only Massachusetts, Maine, and Rhode Island (Table 66). Note that in Massachusetts employment in the water excursion sector has been declining, which means that the Massachusetts share of sector employment has also been declining, hence the decline in the location quotient. If water excursion employment continues to decline in Massachusetts, the location quotient would likely fall below one as long as regionwide employment remains constant or increases.
The Northeast region Herfindahl index was low, ranging from 0.019 in 1998 to 0.013 in 2000. The low value of the index is indicative of an industry with a large number of establishments not dominated by a small number of large establishments. The index value declined by an average of 1% per year from 1998-2005 (Table 67). This decline was primarily due to increasing numbers of establishments although the size structure of the regionwide sector also became slightly more dispersed (i.e., a slight reduction in the proportion of employment in the largest size class in the sector).
Among states the Herfindahl index was generally higher than that of the region as a whole because of a smaller number of establishments. The Herfindahl index declined, on average, in Connecticut, Delaware, Massachusetts, Maine, North Carolina, and Rhode Island.
The location quotient exceeded one in only 27 of the 143 coastal counties included in the analysis. The combined totals for these counties accounted for just over 50% of total Northeast region establishments (Table 68) but accounted for more than 70% of water excursion employment (Table 69). This indicates that the water excursion industry is more geographically concentrated than are other industries in the marine recreational boating subgroup. Note that of the counties listed in Table 68 and Table 69 three (Suffolk County, MA, New York County, NY, and the City of Norfolk, VA) accounted for at least 40% of regionwide employment from 2003-2005.
Shipping and Shipping-related Industries
Shipping and shipping-related industries include water transportation of passengers and cargo (NAICS 48311), ship building (NAICS 336611), and water transportation services (NAICS 4883). Water transportation services includes handling of marine cargo (NAICS 48832), businesses providing navigation services to shipping, (NAICS 48833), port and harbor operations (NAICS 48831), and other water transportation services (NAICS 48839). However, the number of establishments in each of these four sectors is small, and this results in a substantial number of occasions in which employment data are suppressed. Since employment for suppressed fields had to be estimated, the reliability of employment data reported herein is subject to greater uncertainty. To reduce this uncertainty, only data for the industry group as a whole (NAICS 4883) are reported.
Compared to the other marine sectors previously described, shipping and shipping-related industries tend to be more concentrated in terms of both industrial organization and geography. Not surprisingly, water transportation service industries like cargo handling, navigation services, and port operations tend to be colocated with shipping in states and locations with large shipping facilities. Ship building locations do not necessarily correspond to where the shipping industry is active. However, like shipping, the ship building industry is geographically concentrated with more than 90% of employment in only three states (Connecticut, Maine, and Virginia).
In 1998 there were a total of 994 shipping and shipping-related establishments (Table 70) employing 67,000 people (Table 71). Across states, declines in the number of establishments, particularly in Maine and New York, have been partially offset by increases in states like Rhode Island, Maryland, and Virginia. The net effect has been a gradual declining trend in the regionwide number of establishments from a peak of 988 in 2000 to 930 establishments in 2005.
Total mid-March employment in the shipping and shipping-related sector was highest in 1998 but has fluctuated without any notable trend between 61,000–64,000 employees ever since (Table 71). Primarily because of the size of the ship building industry, Virginia employment was nearly 25,000 in all years, representing about 41% of regionwide shipping and shipping-related employment (Table 72). In Connecticut, after experiencing employment declines from 1998-2002, shipping and shipping-related employment increased from 7,800 employees in 2002 to a high of 9,500 in 2005. By contrast, employment in Maine has been on a declining trend from nearly 9,000 in 1999 to 6,300 in 2005.
Employment in shipping and shipping-related industries indicates a degree of specialization across states. For example, ship building represented more than 80% of total industry group employment in Connecticut, Maine, and Virginia and was a substantial majority of employment in Rhode Island (Table 73). By contrast, water transportation and water transportation services were the dominant sectors in Massachusetts, Delaware, New Jersey, and New York. Notably each of these states has one or more major ports that are regional hubs for loading and offloading ocean-going freight and passengers.
Deep sea and coastal water transportation NAICS 48311
Establishments engaged in deep sea and coastal water transportation include conveyance of freight or passengers between foreign or domestic ports. These establishments may provide deep sea transportation or may operate solely in coastal waters. Establishments operating in inland lakes or rivers are excluded as are floating casinos and scenic or sightseeing water transportation. The distinguishing features of establishments classified in NAICS 48311 are that they primarily operate on the open ocean or in coastal bays and the point of departure differs from their destination.
The total number of deep sea and coastal water transportation establishments ranged from a high of 375 in 2000 to 342 in 2004 (Table 74). On average, the annual change in the number of Northeast region establishments has declined by 0.4%. Across states, there were nearly twice as many water transportation establishments in New York than in any other state although the number of establishments declined by an average of 2.1%. The average annual change in establishments was also negative in Connecticut, Maine, and North Carolina. By contrast, the average annual change in number of water transportation establishments was positive in Delaware, Massachusetts, New Hampshire, Rhode Island, New Jersey, and Virginia.
Northeast region annual employment in water transportation establishments was more variable, ranging from a low of 8,600 employees in 2000 to more than 10,000 in both 1998 and 2003 (Table 75). On average the annual change in employment (-1.7%) was higher than the annual change in establishments; this difference suggests an industry where employment per establishment has been declining. Across states employment in New York, New Jersey, Virginia, and Massachusetts accounted for about 80% of regionwide employment. Among these states New York had the lowest numbers of employees per establishment; this ratio suggests an industry size structure consisting of a large number of smaller establishments. As will be shown below, this difference in industry size structure is reflected in the Herfindahl index.
The location quotient for the deep sea and coastal water transportation sector exceeded one in all years only in Virginia and New Jersey (Table 76). The location quotient exceeded one in Connecticut in all years expect 2002 and has exceeded one in Massachusetts in every year since 2001. Thus, there is some evidence of an increase in the relative importance of the water transportation employment share in Massachusetts. By contrast the location quotient in New York fell below one in every year since 2003; this pattern suggests a decline in the relative importance of the water transportation industry in the New York coastal economy.
The Northeast region Herfindahl index ranged between 0.04-0.02 from 1998-2005 (Table 77). The low level of the index indicates a diversified industry in terms of employment size that is not dominated by a small number of large employers. The average annual change in the Herfindahl index was negative; this change indicates a decline in concentration, or equivalently, an increase in diversity in employment size class.
As noted previously, the Herfindahl index for the deep sea and coastal water transportation sector in New York was lower than any other state. Since the Herfindahl index is a function of numbers of entities and employment size class, the low index value for New York is because of the large number of establishments and a low value of the dispersion effect. That is, the water transportation sector in New York consists of a large number of small establishments. By contrast the water transportation sector in Massachusetts exhibits a Herfindahl index ranging from 0.18-0.35. Given the relative stability in numbers of establishments and a fewness effect (1/N) of 0.03, the dominant effect behind the Massachusetts Herfindahl index is the dispersion effect which averaged nearly 0.26. Thus, the magnitude of the Herfindahl index in Massachusetts is primarily due to an establishment size distribution skewed toward fewer large employers. The same may be said of the water transportation industry structure in Virginia although the Herfindahl index has been declining. In Virginia’s case the decline has been primarily due to a decline in the standardized dispersion, indicating a trend toward smaller establishment size in terms of employment.
Just as the industry group as a whole exhibited clear patterns of regional specialization, the deep sea and coastal water transportation sector tended to be concentrated in a small number of Northeast region coastal counties. The location quotient exceeded one in 23 of the 143 coastal counties in the Northeast region. The number of establishments in these counties ranged from a single establishment in Pasquotank County, NC, to more than 100 in Cecil County, MD (Table 78). In terms of employment, the number of employees was highest in Norfolk County, VA, averaging just over 1,200 people (Table 79). Employment in the remaining counties ranged widely from fewer than ten to several hundred.
Ship building and repairing NAICS 336611
The ship building and repair sector comprises establishments engaged in the fabrication or repair of watercraft intended for commercial use. This industry includes shipyards that may be dedicated to the construction of very large ocean-going vessels such as cargo or tankers or the construction of smaller vessels such as commercial fishing or lobster boats. The ship building industry accounts for a comparatively small number of Northeast region establishments but employs a large number of people through a small number of large establishments primarily in Connecticut, Maine, and Virginia. Regionwide, the number of establishments was 139 in 1986 and 142 in 2005 (Table 80). On average the annual change in number of establishment was positive (0.5%) from 1986-1996, but was effectively zero from 1997-2005 although the number of establishments has been trending downward since peaking at 156 in 1999.
Across states the number of establishments ranges from one or two in Delaware and in New Hampshire to more than 30 in Virginia. On average, the annual change in numbers of shipyards was positive for most states, but was negative in Massachusetts, North Carolina, New Jersey, and New York.
Total employment in the ship building sector exceeded 60,000 from 1986-1992 but has since declined to less than 40,000 employees in every year since 2000 (Table 81). Ship building exhibits substantial regional concentration in terms of employment with approximately 91% of regionwide employment in three states; Maine (17%), Connecticut (23%), and Virginia (51%). Much of the regional decline in ship building employment has been due to reduced employment in these three states, although employment has stabilized at around 20,000 in Virginia and has been increasing since 2002 in Connecticut.
Consistent with the regional dominance in ship building employment, the location quotient exceeded one in only Maine, Connecticut, and Virginia (Table 82). In relative terms, this means that the share of ship building employment in these states was substantially greater than the average employment share. Thus, a change in ship building employment in these states would have a proportionally greater effect on coastal economies as compared to other states in the Northeast region.
The ship building Herfindahl index for the Northeast region ranged from 0.152 in 1988 and 1989 to 0.216 in 2004 (Table 83). The index value was the highest among all other marine sectors identified herein and has been increasing over time. The magnitude of the ship building Herfindahl index is due to the dominance of a small number of large employers. The Herfindahl index exceeded 0.9 in both Connecticut and Maine because there was a single employer in each of these states that employed more than 5,000 people. Since 1998, no other establishment in either state employed more than 100. The Herfindahl index was also quite high in Virginia, but the state was more diverse in terms of the range of employment size class compared to either Maine or Connecticut.
Among Northeast region coastal counties, the ship building location quotient exceeded one in 19 of 143 counties. These counties include major ship building facilities in New London County, CT, Sagahadoc County, ME, and in Newport News City, VA. In 2005, these three counties alone accounted for 42% of the regionwide total number of establishments (Table 84) and 83% of total Northeast region employment (Table 85).
Water transportation service NAICS 4883
Establishments in the water transportation services include a wide range of activities that facilitate the flow of marine cargo and passengers from one port to another. These activities include port and harbor operations (docking and pier facilities as well as waterfront terminals and canals), marine cargo handling (stevedoring and lumping services), navigation services to shipping (piloting, tugboats, and ship traffic reporting), as well as floating drydocks, marine surveyors, and cargo checkers. The number of water transportation service establishments was 481 in 1998 but has been declining over time to 437 establishments in 2005 (Table 86).
Total employment in water transportation services grew from 15,048 in 1998 to 16,773 in 2005 (Table 87). Note that employment has been increasing even as the number of establishments has been falling. Employment in this sector was concentrated in the mid-Atlantic states of New York, New Jersey, Maryland, and Virginia. These four states accounted for an average of 85% of Northeast region water transportation services employment. The average annual change in employment was positive in all but three states: Massachusetts, North Carolina, and Rhode Island.
The location quotient exceeded one in all years in Delaware, Maryland, New Jersey, and Virginia (Table 88). Note that these states form a contiguous area from Virginia to New Jersey where the share of employment in water transportation services is greater than both the employment share for the Northeast region as a whole and the average employment share within each state.
The Northeast region water transportation services Herfindahl index was 0.019 in 1998 but has increased to 0.022 in 2005 (Table 89). The higher index in 2005 was due to a small increase in the dispersion effect which is suggestive of a slight shift toward greater concentration of employment in large establishments at least for the region as a whole. Note that this trend in the Herfindahl index is reflected in the overall downward trend in number of establishments coupled with the upward trend in total employment.
This does not mean that any such concentration has been significant since the regional Herfindahl index is still quite low. Among states the Herfindahl was highest in New Hampshire, Connecticut, and Delaware. The New Hampshire Herfindahl index was high because of the small number of establishments (the fewness effect) as none of the establishments in New Hampshire employed more than 19 people. By contrast, the Herfindahl index in Connecticut and Delaware were comparatively high because proportionally more people were employed by a small number of larger establishments.
Among counties the water transportation services sector location quotient exceeded one in 22 of 143 Northeast region coastal counties. Many of these counties include a major port and accounted for 43% of the total number of water transportation service establishments (Table 90) and 84% of regional employment (Table 91).
ENDNOTES
-
Industrial classification systems are used by governments to categorize individual establishments based on similarities in products and/or production technologies for purposes of tracking trends in industries. The Standard Industrial Classification was the classification system in use in the United States until it was replaced in 1997 by the North American Industry Classification System (NAICS).
-
The North American Industry Classification System codes were developed by the Office of Management and Budget in cooperation with Statistics Canada, and the Instituto Nacional de Estadística y Geografía of Mexico.
-
REFERENCES CITED
Donahue Institute. 2006. An assessment of the coastal and marine economies of Massachusetts. Report 1, RFR#:ENV 06 CZM 09. Amherst (MA): University of Massachusetts.
Georgiana D. 2000. The Massachusetts marine economy. Dartmouth (MA): University of Massachusetts Donahue Institute, Center for Policy Analysis.
Hoaglund P, Di Jin, Thunberg E, Steinback S. 2005. Economic activity associated with the northeast continental shelf large marine ecosystem: an input-output approach. In: Sustaining Large Marine Ecosystems: the human dimension. Hennessey T, and Sutinen J Editors. Amsterdam (The Netherlands):Elsevier. 157-179.
Kelly W A 1981. A generalized interpretation of the Herfindahl index. Southern Economic Journal. 48(1):50-57.
Minnesota IMPLAN Group. 2000. Data guide, IMPLAN professional 2.0. Stillwater (MN): Minnesota IMPLAN Group Inc.
Pontecorvo G, Wilkinson M, Holdowsky M. 1980. Contribution of the ocean sector to the United States economy. Science. 208:1000-1006.
Steinback S, Thunberg E. 2006. Northeast regional commercial fishing input-output model. NOAA Technical Memorandum NMFS-NE-188. Woods Hole (MA): National Marine Fisheries Service.
US Bureau of Economic Analysis. 2009. State annual personal income [Internet]. Regional Economic Accounts. [cited 2009 March 23] Available from: http://www.bea.gov/regional/spi/default.cfm?satable=SA04&series=ancillary
US Census Bureau. 2007. County business patterns [Internet]. Economic Planning and Coordination Division, Register Analysis Branch. [cited 2007 September 13] Available from: http://www.census.gov/epcd/cbp/view/intro.html.
US Census Bureau. 2008. County business patterns download page [Internet]. Economic Planning and Coordination Division, Register Analysis Branch. [cited 2008 June 27] Available from: http://www.census.gov/epcd/cbp/download/cbpdownload.html.
