Stynes Chang and Propst 1996 National CE Estimates 02/16/98 Page 1. National Economic Impacts of CE Recreation Visitor Spending: An Update for 1996

Transcription

1 Stynes Chang and Propst 1996 National CE Estimates 02/16/98 Page 1 National Economic Impacts of CE Recreation Visitor Spending: An Update for 1996 Daniel J. Stynes, Wen-Huei Chang and Dennis B. Propst This report updates estimates of the economic impact of Corps of Engineers (CE) recreation visitor spending on the national economy. Previous reports have estimated national economic impacts of CE visitor spending for 1991 (Jackson et. al. 1994) and 1994 (Jackson et. al. 1996) using methods developed by Stynes and Propst (1992). In this report we update the estimates using 1996 visitation data and a 1993 model of the U.S. economy. Some refinements are made to the multipliers and estimates of job impacts in the national models. The new national estimates are a byproduct of a recent study that estimated the economic impacts of visitor spending on local regions surrounding each of the CE's 456 projects. Estimates of spending and impacts for each project for 1996 were accumulated to the District, Division and National levels. A total of $5.6 billion dollars was spent within 30 miles of CE projects on 376 million recreation visits in This spending had a combined local impact (including local multiplier effects) of $6 billion in sales, $3.2 billion in income and 180,000 jobs (Propst, et. al. 1998). The national impact of visitor spending, however, is much more than the sum of these local impacts. For a national estimate of impacts, another $3.4 billion in trip spending outside the 30 mile radius is included, and a national model is used to estimate capture rates and multiplier effects. The national economic effects of visitor spending on recreation trips to CE projects in 1996 were $24.5 billion in sales, $12.3 billion in income and 420,000 jobs. The national impact is roughly four times the combined "local" impacts in terms of sales and income and 2.6 times the number of jobs from just the local effects. The revised national estimates benefit from several refinements in the procedures, while still being sensitive to the usual errors in estimates of visitation, average spending, and multipliers. Methods Visitor spending is estimated using the same segmented procedures as in earlier reports. Visitation and spending are estimated for six types of visitors: day users, campers, and other overnight visitors with each group split between boaters and non-boaters. As in previous studies, some difficulties were encountered in estimating the percentage of visitors in each of these segments. In particular, distinct estimates of the number of campers are obtained depending on whether camping revenue data is used or manager estimates of the percentage of visitors who are camping (from CE NRMS database). Manager estimates of camping activity are almost four times larger than those derived from the revenue data. A revenue-based estimate of 7.6 million camper visits on CE projects is assumed to be more accurate, although several assumptions were needed to convert camping revenues to visits. These assumptions are discussed more fully in Propst et. al. (1998). Camping's share of overall visitation will affect the spending estimates as campers spend considerably more than day users on a per visit basis. Also, there are only crude estimates of the percentage of CE visitors staying overnight in lodges, cabins and motels on the project or in campgrounds, motels or other lodging off the project. These visitors are all counted as day users in the CE visitation data, although they have quite different spending patterns. Lacking any reliable data on these "other overnight" visitors, the same assumption was made as for the 1994 estimate. One percent of day use visits were assumed to be "other overnight" visitors, staying overnight in the area and having a higher rate of

2 Stynes Chang and Propst 1996 National CE Estimates 02/16/98 Page 2 spending than day users. Other overnight visitors spend about $100 per party per day in the local area, compared to about $40 per day for day users. As the average stay for other overnight visitors was about 3 days, they represent 3% of all visitors on a party-day basis. Spending profiles for each segment were taken from the same source as earlier national reports, surveys conducted by Propst, Stynes, and Lee (1992) at 12 CE projects in 1989/1990. The spending averages were converted to a party-day basis by dividing spending per party per trip by an average length of stay for each segment. The spending profiles were also price adjusted to 1996 using Bureau of Labor Statistics (BLS) consumer price indices for each major category of spending. The number of visits in party days and average spending per party day are reported in Table 1. Day users constitute 92.5% of party days nationally at CE sites, while campers make up 4.5% and other overnight visitors about 3%. Spending per day for the entire trip varied from a low of $53 for day users who are not boating to $160 per party per day for other overnight visitors who are boating. Campers spent about $100 per day, $96 for boaters and $114 for non-boaters. Campers who were boating spent more ($66) than non-boaters ($50) within 30 miles of the site. Campers who were not boating spent considerably more outside the local area. Overall, 62 percent of the $9 billion in total visitor spending ($5.6 billion) occurs within 30 miles of CE projects and generates local impacts, while 38 percent or $3.4 billion occurs more than 30 miles from projects. Table 1. Summary of Total CE Visits and Spending a, 1996 Camper Day User Other Overnight Boat Non-Boat Boat Non-Boat Boat Non-Boat Total Visits ( Party-Days, MM) Percent of days by segment 1.0% 3.5% 19.9% 72.7% 0.7% 2.2% 100% Average spending (dollars per party-day) Within 30 miles of project Entire trip Percent within 30 miles 69% 44% 64% 63% 81% 63% 62% Total Spending ($million) Total Spending within 30 miles ($MM) ,505 3, ,596 Total Spending for Entire Trip ($MM) ,337 5, ,967 Percent of Total by segment 1.5% 6.2% 26.1% 59.8% 1.6% 4.7% 100% a. All spending is in 1996 dollars. Days users who are not boating make up 73 percent of all CE visitors on a party day basis and contribute 60 percent of overall spending. Day users who are boating at CE projects represent twenty percent of visitors and 26 percent of spending. Campers are 4.5 percent of visitors, but account for almost eight percent of all spending. Other overnight visitors are about three percent of visitors but spend over six percent of the $9 billion in total spending. Economic impacts of CE visitor spending are estimated by applying the estimates of visitor trip spending to an input-output (I-O) model of the U.S. economy. The I-O model is estimated using the IMPLAN system with the

3 Stynes Chang and Propst 1996 National CE Estimates 02/16/98 Page databases. The DOS version of IMPLAN is used, but the Type III multipliers are adjusted downward to correct for a bias in the way that IMPLAN computes these multipliers. IMPLAN's simulates induced effects by estimating household formation based on the direct and indirect job impacts. It then applies an average household income to determine the amount of income to re-circulate as induced effects. The resulting Type III multipliers are biased upward for tourism applications, as tourism jobs involve about half the income of an average job. This is due both to lower wages and a higher percentage of part time and seasonal jobs. The result is that the IMPLAN Type III multipliers overestimate the induced effects of tourism spending by a factor of about two. Halving the induced effects yields an aggregate multiplier similar to the IMPLAN Type II multiplier (only available in the Windows version) and to multipliers produced by the Bureau of Economic Analysis's RIMS II model (Regional Input-Output Modeling System). Estimates of secondary and total effects in our previous reports are biased upward, as this correction was not made. The national multipliers for recreation spending are reported in Table 2. Type I multipliers include direct and indirect effects, while Type III multipliers also include induced effects. Ninety-five percent of visitor spending is captured by the national economy, as about five percent of visitor spending goes for imported goods. Every dollar of direct sales generates another 63 cents in indirect effects at the national level. The adjusted national Type III sales multiplier is 3.15, which means that total sales effects include 63 cents of indirect sales and $1.52 of induced sales for every dollar of direct sales. A Type III sales multiplier of 4.26 was used in the 1991 and 1994 studies, which explains the much higher estimates of secondary effects in those reports. About half of all direct sales is converted to income and roughly 19 jobs are supported by each million dollars in direct sales. Including secondary effects, each dollar of direct sales yields $1.58 in total income and each million dollars in direct sales supports 45 jobs (Table 2). The job to sales ratios from the national model are biased downward, as the ratios are substantially higher in regions around CE projects, where the majority of direct jobs occur. We adjust for this bias later in this report. Table 2. Economic Multipliers for The United States a Sales Income b Jobs c Direct effects Type I multiplier Type III multiplier Capture rate d 90% a. Multipliers are from a 1993 input-output model for the United States constructed using IMPLAN b. Income per dollar of direct sales. Income includes employee compensation, proprietor's income and other property income. c. Jobs per million dollars in direct sales d. Capture rate = direct sales/total visitor spending

4 Stynes Chang and Propst 1996 National CE Estimates 02/16/98 Page 4 National Economic Impacts of All Trip Spending CE recreation visitors spent $9 billion dollars on trips to CE projects in This spending resulted in $7.8 billion in direct sales effects, $3.9 billion in direct income and supported 150,000 direct jobs (Table 3). Another $8.7 billion in income was generated through secondary effects, divided 27% to indirect effects and 73% to induced effects. Recreation trip spending accrues mainly to service and retail sectors, which convert a high percentage of sales to income. Visitor spending therefore has fairly high induced effects and smaller indirect effects, compared to manufacturing sales. Of the 350,000 jobs supported by CE visitor spending over 200,000 are from secondary effects. Sales and income are expressed in 1993 dollars as spending is deflated to the model year. Table 3. National Economic Impacts of Recreation Visitor Spending (1996), All Trip Spending a Sales ($millions) Income ($millions) Jobs (number of jobs) Direct Effects 7,783 3, ,461 Indirect Effects 4,940 2,277 46,796 Induced Effects 11,777 6, ,750 Total Effects 24,500 12, ,007 a. Sales and income are in 1993 dollars Figure 1 shows the distribution of income from CE visitor spending across economic sectors. Retail and wholesale trade sectors receive $1.8 billion of the $3.9 billion in direct income from visitors, followed by manufacturing/production sectors ($891 million), and eating and drinking establishments ($522 million). Hotels, amusements, and transportation/services each receive more than $200 million in direct income. The largest shares of secondary income accrue to services, manufacturing and retail trade. Income is the best measure of the contribution of CE visitor spending to gross national product, as it captures the key components of value added to the economy by CE recreation activity. Employment measures, while more frequently cited, tend to be distorted by differences in wages and the proportions of part time and seasonal jobs across sectors. Sales figures can also be misleading as sectors differ in the proportion of sales that is converted to income. For example, manufacturing may generate a high sales volume, but much of the revenue from these sales may go to purchase parts. This creates high levels of indirect effects if the inputs are purchased within the region, but produces less direct income per dollar of sales, when compared to more labor intensive sectors like services and recreation.

5 Stynes Chang and Propst 1996 National CE Estimates 02/16/98 Page 5 Income Effects by Sector 5,000 4,500 4,000 SECONDARY DIRECT 3,500 $Millions 3,000 2,500 2,000 1,500 1, Manf/Prod. Trans & Services Recreation Hotel Eat & drink Retail Govt Figure 1. Distribution of Income Effects by major Sector Groups National Economic Impacts of Spending Near CE Projects By applying only the $5.6 billion in visitor spending that is made within 30 miles of a CE project to the national I-O model, we can estimate the national impacts of spending near CE projects. This local spending results in $4.8 billion in direct sales, $2.4 billion in income and supports about 98,000 local jobs. Sales and income are expressed in 1993 dollars as spending is deflated to the model year. The secondary effects of local spending has impacts throughout the economy. As most CE projects are in rural areas, most of the goods bought by visitors are manufactured outside of the local area. The impacts on manufacturing sectors are therefore felt in other regions of the country. Indirect and induced effects of local spending are also felt well beyond the 30 mile radius of where the spending occurred. Thus, the national impacts of local spending are a mix of mostly local direct effects and more Table 4. National Economic Impacts of Recreation Visitor Spending Within 30 Miles of Projects, 1996 a Sales ($MM) Income ($MM) Jobs (number of jobs) Direct Effects 4,789 2,426 98,137 Indirect Effects 3,123 1,422 29,735 Induced Effects 7,673 3, ,176 Total Effects 15,585 7, ,047 a. Sales and income are in 1993 dollars

6 Stynes Chang and Propst 1996 National CE Estimates 02/16/98 Page 6 widely dispersed secondary effects. With secondary effects, visitor spending within 30 miles of CE projects generates a total impact on the national economy of $15.6 billion in sales, $7.8 billion in income and 228,000 jobs. These impacts are about 64% of the corresponding national impacts when all trip spending is included (i.e., Table 3). The impacts of local spending on the national economy are neither purely national or purely local. They show the impacts on the larger national economy of spending near CE projects. These estimates may appeal to those who feel that spending near the destination is more directly attributable to the recreation opportunities being provided. While necessary trip expenses outside this distance should likely be included in an impact analysis, the wider spending region will also include incidental expenses and spending associated with other stops on multi-stop trips. Comparisons with 1994 National Estimates and the Sum of Local Impacts for 1996 The difference between national and local impacts may be more fully understood by comparing the two national impact estimates from Tables 3 and 4 with a purer "local impact" estimate. This comparison also reveals a bias in the estimates of job impacts from national models. Propst et. al. (1998) estimate the impacts of spending near CE projects in 1996 for each of the 456 CE projects using models/multipliers for the local regions. The sum of these local impacts (the first column of figures in Table 5 ) measures the impact of local spending (within 30 miles of projects) on the local economies. Since some refinements have been made to the impact estimation procedures for 1996 (in particular adjustments of multipliers), we also compare the 1996 national estimates with similarly adjusted 1994 figures in Table 5. The first two columns of figures in Table 5 compare local versus national impacts of spending within 30 miles of CE projects. The middle two columns may be used to compare national impacts of all trip spending versus just spending within 30 miles of projects. Finally, the two right hand columns compare 1994 and 1996 national impacts of all visitor spending. All impact results are presented in 1993 dollars. First, let's compare the two estimates of impacts from visitor spending within 30 miles. On average, only 66% of visitor spending within 30 miles of a CE project is captured by the local economy, while 91% of this spending is captured by the national economy (Table 5 ). When considering the local region, purchases of services and only the retail margins on most goods purchased by visitors accrue to the local economy. Hence, the national economy captures an additional $1.3 billion of the local spending as direct sales and then another $3 billion in secondary effects of these sales. Secondary effects are also substantially larger for the national model than for local regions. The average Type III sales multiplier for local regions around 108 of the 456 CE projects was 1.67, well below the national sales multiplier for local spending of This difference in multiplier effects explains another $5.5 billion in secondary impacts of local spending that are captured by the U.S. economy but not within the local areas around CE projects.

7 Stynes Chang and Propst 1996 National CE Estimates 02/16/98 Page 7 Table 5. Comparison of National and Local Impact Estimates Regional Model Local models U.S. Model U.S. Model U.S. Model Year a Spending covered Within 30 Within 30 Total Trip Total Trip Visits (person trips, MM) Day Users Campers Total Spending in 1996 dollars, except final column which is in 1994 dollars Spending Averages Per party-day Per party-trip Total Spending ($MM) 5,596 5,596 8,966 8,149 All spending, multipliers and Impacts below in 1993 dollars Total Spending ($MM) 5,325 5,257 8,613 8,068 Multipliers Capture Rate 66% 91% 90% 87% sales (total sales to direct sales) income (total income to direct sales) jobs (total jobs to direct sales) Sales Effects ($millions) direct 3,512 4,789 7,783 7,048 secondary 2,337 10,796 16,717 13,597 total 5,851 15,585 24,500 20,646 Income Effects ($millions) direct 1,821 2,426 3,915 3,536 secondary 1,266 5,420 8,413 6,966 total 3,087 7,846 12,328 10,501 Job Effects (number of jobs) b direct 120,906 98, , ,000 secondary 42, , , ,057 total 163, , , ,057 a. Multipliers and secondary effects for 1994 are adjusted for exaggerated induced effects from the original IMPLAN Type III multipliers. b. Job effects from the national models are biased downward. See Table 6 for the adjusted estimates.

8 Stynes Chang and Propst 1996 National CE Estimates 02/16/98 Page 8 In summary, the difference between the $16 billion in national sales effects of spending within 30 miles of CE projects and the $6 billion in purely local impacts of this same spending is roughly $5 billion in additional spending that is captured by firms outside the local area (including associated multiplier effects) and another $5 billion in additional multiplier effects. Including the other $3.4 billion spent outside the local region and the secondary effects of this spending brings the national impact of all trip spending from $16 billion to $25 billion in total sales effects. The differences in income and jobs between columns in Table 5 follow from the differences in sales. Comparing the direct job estimates between the national and local models in the first two columns, however, illustrates a problem with the employment estimates. The national model predicts about 20,000 fewer direct jobs even though it captures $1.3 billion more in direct sales. The problem stems from the difference in job to sales ratios for local regions compared to the national averages. Local regions around CE projects average 34 direct jobs per million in direct sales, while the U.S. as a whole averages only 20 jobs per million (Table 6). As the direct jobs are created primarily in the local areas around CE projects, the local ratio should be more accurate. National job impacts are adjusted for this problem in Table 6. The local models were assumed to yield an accurate estimate of the direct jobs (120,906) from the $3.5 billion in direct local sales. The revised national estimate of job impacts from local spending is 147,065. This figure includes the 120,906 jobs in local regions plus another 26,159 jobs from direct sales that is captured outside the local areas. The national job to sales ratio of 20 jobs per million was applied to the additional $1.3 billion in direct sales to get the 26,159 additional jobs. This same approach was used to adjust the secondary job estimates. The local job to sales ratios are applied to sales captured within the local regions and the national ratios are applied to sales outside this area. The same approach was used to adjust national job impact estimates for all trip spending. National job to sales ratios were applied to the sales associated with spending outside 30 miles and added to the jobs estimated for spending within 30 miles. With these adjustments, the 1996 national employment effects increase to 420,000 jobs from all trip spending and 292,000 jobs from spending within 30 miles of CE projects. The greatest change from the job estimates in Table 5 are for direct jobs as this is where the job to sales ratios differ the most. Table 6. Adjustments to National Job Impact Estimates Regional Model Local models U.S. Model U.S. Model U.S. Model Year Spending covered Within 30 Within 30 Total Trip Total Trip Jobs to Sales ratios (1993 dollars) Direct jobs Secondary jobs Total jobs Adjusted Jobs direct 120, , , ,534 secondary 42, , , ,411 total 163, , , ,946

9 Stynes Chang and Propst 1996 National CE Estimates 02/16/98 Page 9 The final two columns of Tables 5 and 6 are the national impacts of all trip spending for 1994 and All figures are adjusted to 1993 dollars to facilitate comparisons. The 1994 estimates are based on a 1990 IMPLAN model, while the 1996 estimates use a 1993 model. Estimates of secondary effects for 1994 have been adjusted downward from the figures reported in Jackson et. al. (1994) as those are based on the biased IMPLAN Type III multipliers. Corresponding adjustments are also made to the employment effects for 1994 in Table 6. All of these adjustments make the 1996 and 1994 estimates as comparable as possible. Impacts in 1996 are based on a slightly smaller number of visits than in 1994 and a lower percentage of campers (Table 5). Spending grew from $58.15 per trip in 1994 dollars to $67.61 in 1996 dollars. The increase over and above price changes between 1994 and 1996 is due to a change in the percent of day users assumed to come from the local area. The 1994 estimate assumed 80% of day users came from within 30 miles, while the 1996 estimate used 50%. Day users from outside the local area spend more than locals en route. Total sales and income grew by about eighteen percent between 1994 and 1996 (Table 5), while jobs (after adjustment) grew by 12 percent (Table 6). Growth was somewhat higher in secondary effects than in direct effects due to modest increases in multipliers between the 1990 and 1993 national models. Conclusions This report has updated estimates of the national economic impacts of CE visitor trip spending for 1996 and made some refinements to the impact estimation procedures. In 1996, million recreation visitors spent $9 billion on recreation trips to CE projects, spending $5.6 billion within 30 miles of CE projects. Trip spending resulted in $7.8 billion in direct sales, $3.9 billion in direct income and 205,000 direct jobs. Total impacts including secondary effects were $24.5 billion in sales, $12.3 billion in income and 420,000 jobs. These national estimates are compared with the impacts on local regions around CE sites and with national estimates for Using total income as the best measure of the contribution of visitor spending to a region's economy, we find total income effects grow from $3.1 million for the impacts of local spending (within 30 miles of CE projects) on local regions, to $7.8 million in income effects of local spending on the national economy, to $12.3 million in income effects of all trip spending on the U.S. economy. Comparisons between local and national estimates of impacts provides a clearer picture of the distribution of impacts, while also helping to identify a bias in job estimates, when national models are applied to primarily local spending. Estimates of the direct job impacts of CE visitor spending are 50 percent higher than predicted by a national model. The direct job impacts of all trip spending are 37 percent higher. Estimates of total job impacts are therefore underestimated by twenty to thirty percent when national models are used to estimate the number of jobs stemming from sales in rural and tourist-oriented economies. The sensitivity of job estimates to the number of part time and seasonal jobs is another reason for using income as the preferred impact measure in recreation and tourism applications.

10 Stynes Chang and Propst 1996 National CE Estimates 02/16/98 Page 10 References Jackson, R. S., Stynes, D. J., Propst, D. B., & Siverts, L. E Economic impact analysis as a tool in recreation program evaluation. Instructional Report R Department of the Army, Waterways Experiment Station, Vicksburg, MS. Jackson, R.S., Stynes, D.J., Propst, D.B An assessment of the national economic effects of the U.S. Army Corps of Engineers recreation program. Vicksburg, MS: Environmental Lab, Waterways Experiment Station, Department of the Army, Corps of Engineers. 15pp. Jackson,R.S., Stynes,D.J., Propst,D.B., Carlson,B.D National and State Economic Effects of the 1994 U.S. Army Corps of Engineers Recreation Programs.Vicksburg, MS:U.S. Army Engineer Waterways Experiment Station, Technical Report R pp. Propst, D.B., Stynes, D.J., Lee, J.H. & Jackson, R.S Development of spending profiles for recreation visitors to Corps of Engineers projects. (Technical Report R-92-4, 112 pp.). Vicksburg, MS: U.S. Army Engineer, Waterways Experiment Station. (Executive summary also available as Technical Report R-92-1, 16 pp.). Propst, D.B., Stynes, D.J., Chang, W.H., and Jackson, R.S Estimating the local impacts of recreation at 456 Corps of Engineers Projects. East Lansing, MI: Department of Park, Recreation and Tourism Resources, Michigan State University. Final Report to U.S. Army Corps of Engineers, Waterways Experiment Station. Stynes, D.J. and D.B. Propst A system for estimating local economic impacts of recreation and tourism. In. Measuring tourism impacts at the community level. S. Reiling (Ed). Maine Agr. Expmt. Sta. Misc. Report #374.