State-Level Economic Contributions of Active Outdoor Recreation Technical Report on Methods and Findings

State-Level Economic Contributions of Active Outdoor Recreation Technical Report on Methods and Findings April 13, 2007 Prepared by Southwick Associates, Inc. Fernandina Beach, Florida For: Outdoor Industry Foundation i

Acknowledgements This study was conceived and funded by the Outdoor Industry Foundation. Data collection was contracted to Harris Interactive, Inc. of Rochester, NY. and economic research and report development was contracted to Southwick Associates, Inc. of Fernandina Beach, FL. The primary authors were Rob Southwick and John Bergstrom, Ph.D. The authors wish to express their desire to thank Clint Wall, OIF s project manager, for his direction, assistance and support. Many people provided time and thought towards the success of this project. The authors and OIF wish to thank many people, including Carolyn McKernan of REI, Inc., for her input and recommendations, Courtney Blacker of Jansport, Colleen Pate of Coleman, John Schoene of Johnson Outdoors, and many other contributors. Despite the contributions of many, the authors remain responsible for the contents herein. ii

Table of Contents 1.0 Introduction 1 2.0 Definitions 2 3.0 Methodology 4 3.1 Categories under Study 4 3.2 Study States 5 3.3 Methods Overview 6 3.4 Primary Data Source Harris Interactive Survey 7 3.5 Estimating the Number of Participants and Trips 9 3.6 Estimating Expenditures 15 3.6.1 Equipment & Services Expenditures 15 3.6.2 Trip-Related Expenditures 16 Sidebar: The Size and Economic Effects of Outdoor Swimming 23 3.7 Calculating Economic Contributions 23 Sidebar: The Magnitude of Trip Impacts versus Equipment Impacts 28 3.8 Estimating Wildlife-Based Recreation Contributions 28 3.9 The Combined Effects of All Outdoor Recreation 32 3.10 Taxes 36 3.11 State-Specific Results Versus Regional Summaries 36 Bibliography 37 Appendix I: When State-Level Expenditure Items Were Substituted with the National Average (State Value >25% Greater Than the National) 38 Appendix II: Detailed Results: Total Expenditure, Economic Contribution, Trip and Participation Estimates for Active Outdoor Recreation, Per State 59 Appendix III: Detailed State-Level Wildlife-Based Recreation Impacts 102 Appendix IV: Survey Statistics 145 Appendix V: Survey Questionnaire 148 iii

1.0 Introduction This report assesses the economic contributions of the major forms of active outdoor recreation for 21 states. Activities included are: Bicycle-based recreation (on paved roads or off-road) Camp-based recreation (Camping in an RV at a campsite, in a tent, or rustic lodge)1 Paddle-based recreation (kayaking (recreational/sea/whitewater), rafting and canoeing) Fishing (all types) Hunting (all types) Snow-based recreation (cross-country/nordic skiing, downhill skiing, telemark skiing, snowboarding and snowshoeing Trail-based recreation (trail running (running on an unpaved trail), day hiking on an unpaved trail, backpacking and climbing ice or natural rock) Wildlife viewing (wildlife watching and birding) The purpose of this project is to help readers understand the relative and overall economic significance of recreation to 21 state economies. This data and text within this report are derived from a similar 2006 OIF report by the same authors that examined the identical impacts at the regional and national levels. This report is based on data from a Harris Interactive survey commissioned by the OIF in late 2005 covering bicycle, camp, paddle, snow, and trail-based recreation; and existing data from the 2001 National Survey of Fishing, Hunting and Wildlife-Associated Recreation commissioned by the U.S. Fish and Wildlife Service. Economic impact estimates were constructed using well-established modeling procedures. Existing scientific publications and reports were used to ground truth to the final results. Procedures used and data limitations encountered are described in this report. This study developed economic impact estimates for the major forms of active outdoor recreation of interest to OIF. By definition, these exclude categories that rely on the primary use of internal-combustion engines such as motorsports (off-road), boating, snowmobiling, etc. While the purchases of vehicles were discounted from this study by treating them as outliers, trip costs associated with vehicles such as fuel were not excluded. 1 Purchases for RVs and other motorized vehicles were discounted. Operating costs were included in this study but not the purchase price. This adjustment is explained later further in this report. 1

2.0 Definitions Economic impacts measure the changes within an economy. For the purposes of this report, economic impact refers to the growth or contraction in an economy caused by the entry or loss of revenue from outside sources. Outside sources include tourists bringing dollars into a local economy or the sale of services and products to people outside of the region. Sales and other transactions between people and businesses within an economy typically dos not result in economic growth but is mainly a redistribution of resources. However, this form of economic activity is still important and sustains jobs and more. This form of economic activity will be referred to as economic contributions within this report, and includes both impacts from recreationists visiting from outside the economies under study and the economic contributions from residents within the study areas. Economic contributions are usually expressed as jobs, income, retail sales (expenditures) and tax revenues. Economic contributions and impacts, for the purpose of economic modeling, can be divided into three standard components: direct, indirect and induced effects. The indirect and induced effects are the two components of the multiplier or ripple effect. Each of these is considered by most economic models when estimating the overall effects of any activity on the economy. A direct effect is defined as the result of the initial purchase made by the consumer. Only the amount of the purchase that remains in the region under study is retained as the direct effect. For example, when a person buys a restaurant meal for $20, there is a direct effect to the restaurant and the local economy, of $20, assuming all of the supplies needed for the meal were provided locally. However, recognizing much of the consumed food and supplies were likely bought from sources outside of the region of study, a lower amount, for example, $10, actually remains in the local economy as a direct effect. Indirect effects measure how sales in one industry affect the various other industries providing supplies and support. For example, the restaurateur must purchase additional food and supplies, plus pay costs such as power, rent, etc.; local food suppliers must buy more products, and so on. Therefore, the original direct effect of $10 benefits many other industries within the regions. An induced effect results from the wages and salaries paid by the directly and indirectly impacted industries. The employees of these industries then spend their incomes. These expenditures are induced effects which, in turn, create a continual cycle of indirect and induced effects. The sum of the direct, indirect and induced effects is the total economic impact or contribution. As the original retail purchase (direct effect) goes through round after round of indirect and induced effects, the economic contribution of the original purchase is multiplied, benefiting many industries and individuals. Likewise, the reverse is true. If a particular item or industry is removed, the economic loss is greater than the original retail sale. Once the original retail purchase is made, each successive round of spending is smaller than the previous round. When the economic benefits are no longer measurable, the economic examination ends. There is one economic measure not considered in this report, but commonly used in resource allocation debates: economic value. Economic value measures the personal or intrinsic value held by users of a resource, or people affected by an action or item. This term basically 2

measures the quality of life effects, or how much one is better or worse off intrinsically. For example, a person may spend $100 to go on a raft trip, but the trip was worth $125 intrinsically to that individual. That person was better off by $25 after taking the trip, and $25 is the net economic value of the trip. While an important measure, economic value is complex and not considered in this report. 3

3.0 Methodology 3.1 Categories under Study The Harris Interactive survey examined five categories of active outdoor recreation, most of which can be further divided. Survey recipients were queried about their annual participation and expenditure levels for each of these categories: Bicycle-based recreation: paved roads off-road Camp-based recreation2: recreational vehicles campgrounds/campsite rustic lodges Paddle-based recreation: kayaking (recreational/sea/whitewater) rafting canoeing outdoor swimming (lakes, rivers, natural water bodies) outdoor pool swimming3 Snow-based recreation: cross-country/nordic skiing downhill skiing telemark skiing snowboarding snowshoeing Trail-based recreation: trail running (running on an unpaved trail) day hiking on an unpaved trail backpacking climbing (ice or natural rock) Other forms of active outdoor recreation were excluded from this study. By definition, these categories could have been included, but the cost of identifying enough people in the survey who recently participated in these categories was considered cost prohibitive. This study focused on the major sports within the OIF s primary missions. The exclusion of these other recreations, including categories such as SCUBA, snorkeling, surfing, wind- and kite-surfing, 2 Camping includes day trips when camping equipment was used. 3 Outdoor swimming and outdoor pool swimming were eliminated during an interim round of analysis based on industry feedback that these activities did not fully fit as an active outdoor recreation, plus their sheer size overshadowed the impact of the other activities. More details are presented in the methodology discussions. 4

introduces a level of conservatism to the final results when reporting the economic impacts of active outdoor recreation. This study also included wildlife based recreation: Hunting; Fishing; and Wildlife Viewing. Wildlife-based recreations were handled differently in this study. Information on the economic contributions from these categories were already available from other sources, and only needed updating to 2005 levels to be included in this project. 4 Details are presented in the methodology discussions. As a first time study, there were many decisions to be made about methodology. When these decisions presented themselves, efforts were made to err on the side of conservatism. 3.2 Study States Survey resources did not permit the development of state-level impacts for all states; therefore states were divided into primary and non-primary states. Primary states were designated by the OIF based on educational and communication needs. To permit reliable and reportable economic data, 125 completed surveys were allocated to each primary state. The primary states initially designated were: Alaska Arizona California 5 Colorado Idaho Louisiana Massachusetts Maine Montana North Carolina New Hampshire New Mexico Nevada New York Oregon Tennessee Utah Vermont Washington Wisconsin West Virginia Wyoming The 28 remaining states, designated as non-primary states, were allocated 50 completed surveys each. To permit reportable economic data, these states were initially combined into 4 These sources were: a. American Sportfishing Association. 2002. Sportfishing in America, Values of Our Traditional Pastime. Alexandria, VA. b. International Association of Fish and Wildlife Agencies. The Economic Contributions of Hunting, 2001. Washington, D.C. c. U.S. Fish and Wildlife Service. The 2001 National and State Economic Impacts of Wildlife Viewing. Arlington, VA. 5 California was originally combined with Hawaii, but very few survey responses were obtained for Hawaii. The effects of the Hawaii respondents were removed from the data to produce estimates for California only. 5

eight regions based on shared characteristics and proximity. 6 Based on the limited sample sizes, these states are not used in the state-specific results. 3.3 Methods Overview The economic estimates in this report are based on the expenditures made for recreation within each state. The following standard formula presents the general approach used to determine state expenditures: State Expenditures = (Number of trips in the state x Average expenditures per trip) + Annual equipment expenditures for all participants in the state The above formula has three major data requirements: a) Annual trips for each category for each state; b) Typical expenditures per trip, for each category in each state; and c) Annual expenditures for equipment, by participants in each category in each state. A search for reliable data, or existing literature, for each of the three major requirements did not turn up any reliable sources that provided uniform data across all states and for each category. Therefore, new data had to be generated. To meet this need, in late 2005 the OIF contracted with Harris Interactive to conduct an online survey of outdoor recreationists. The analysis was undertaken in four major stages: 1) Develop estimates regarding the number of people participating in each of the active outdoor recreations, plus the number of trips taken annually and the average amount spent per trip; 2) Develop estimates of the total annual expenditures made by recreationists at the state levels for each category examined; 3) Calculate annual state-level economic impacts for each category using accepted modeling procedures; and 4) For wildlife-based recreations (hunting, fishing and wildlife-viewing), update estimates already available from existing sources. The next sections describe the primary data collected, how data were collected and the steps taken to produce the final economic impact estimates. 3.4 Primary Data Source Harris Interactive Outdoor Survey Harris Interactive, a nationally recognized polling and surveying firm, has developed a proprietary data base of U.S. residents who agreed to participate. Using this database, in October of 2005, Harris Interactive conducted 125 online interviews of qualified residents in each of the 21 primary states. (Harris failed to collect an adequate sample size to create state specific data for Massachusetts.) A qualified resident, or qualified respondent, is a state resident 6 After the first draft results were developed, the regions were adjusted based on feedback from industry members and sample size limitations. This adjustment will be presented later in the methodologies section. 6

who participated in the past year in at least one of the five recreations under study and spent money on at least one of the categories. A total of 5,150 surveys of qualified respondents were completed. In addition to the qualified residents, an additional 8,756 surveys were completed by panel participants who did not participate in any of the targeted recreations or spend money for these categories. These participants, known as non-qualified respondents, permitted the development of estimates of the percent of the U.S. population and state populations that participated in active outdoor recreation in the past year. These participants were not used to develop expenditure estimates. In total, 13,907 completed interviews were obtained. Table A presents state sample sizes. state. Table A: Sample Sizes Participation Sample Expenditure Sample "Non-Qualified" "Qualified" ALASKA 268 105 ARIZONA 340 89 CALIFORNIA* 420 76 COLORADO 408 85 IDAHO 297 15 LOUISIANA 479 83 MAINE 355 89 MONTANA 349 89 NEVADA 485 72 NEW HAMPSHIRE 370 86 NEW MEICO 431 88 NEW YORK 478 82 NORTH CAROLINA 412 77 OREGON 337 85 TENNESSEE 417 85 UTAH 368 74 VERMONT 313 92 WASHINGTON 370 76 WEST VIRGINIA 346 83 WISCONSIN 325 83 WYOMING 269 98 * California was combined with Hawaii in the original Harris survey. Considering only three from the Qualified sample were from Hawaii, the total number of Hawaiians in the Non-Qualified set is expected to be minimal. The CA-HI data was split based on the percentage of the region's residents residing in each state. Weighting: Most samples of residents do not accurately represent the general U.S. population. A weighting process was applied by Harris Interactive to adjust the results to represent the U.S. population 18 years and older. Weights are statistical percentages used to increase or decrease a respondent s survey response to reflect a portion of the population under study. When all survey responses are weighted and summed, the results reflect the entire population under study. Harris applied two weights: one for demographics and the other reflecting the propensity of people fitting the respondent s profile to complete online surveys. The 7

demographic weights, developed by Harris Interactive using their undisclosed proprietary processes, were based on variables, such as race, income, education, age, and gender were used to adjust the sample set to better reflect the true make-up of each state s population. When asked if the sample set adequately reflects the percentage of households with children, Harris Interactive replied in the affirmative. This was asked based on the survey questions about expenditures made for family members under 18 years of age. The propensity weights were based on Harris s experience with online surveys and used to adjust for the respondent s likelihood of participating in online surveys and panels. These characteristics are not necessarily based on demographic variables but also include attitude and behavioral characteristics. Statistics regarding sample sizes and confidence intervals for expenditures are presented in Appendix IV. Assumptions and Caveats to the Harris Interactive Data: 1) A source of conservatism may be introduced into the final economic impact results by the survey questionnaire. The questionnaire, presented in Appendix V, is very specific in the categories included and not included in this project. Some respondents may have excluded expenditures that were justified for inclusion. For example, the bicycle category is meant to include all forms of recreational cycling. This category is divided into paved road and off-road cycling. Survey respondents may use urban trails, which are growing across the U.S., as a primary place to cycle, but may not consider these venues to be either paved roads or off-road and therefore may exclude such cycling. The level of effects from this possible issue is unknown. 2) An unknown level of recall error may have occurred. This occurs when respondents are asked to report "typical" trip expense but instead report a higher expenditure for a trip they fondly recall. 3) As with any survey, there is some level of natural error associated with not finding people who participated in an activity or purchased specific items. This is common for activities and items of lesser popularity. Examples include non-resident trips to Alaska for snow sports, and people buying rarer items such as souvenirs as part of cross-country ski trips. Larger sample sizes tend to diminish this problem, but even with major surveys such as the multi-million U.S. Fish and Wildlife Service s 2001 National Survey on Fishing, Hunting and Wildlife-Associated Recreation, this problem still occurs. This problem often results in low numbers of estimated trips, lower than what may really be occurring, which leads to underestimated expenditures and economic impacts. Appendix IV presents the errors associated with the trip estimates which can be used to identify those estimates with a smaller or greater range of error. As noted elsewhere in this document, when the average expenditure for a state for a specific item was 25% or greater than the national average, it was replaced by the national average to help reduce the chance of overestimating impacts. When the average was 25% or less, it was left as-is to maintain this report s conservative nature. 3.5 Estimating the Number of Participants and Trips 8

NOTE: The OIF has produced dual sets of participation data, using two different survey methods and produced for different purposes. One set of participation data was produced as part of this project and presented in this report. The other set was produced as part of a different effort and not discussed further in this report. When comparing expenditures and participation, please use the participation data presented in this report to ensure proper comparisons. Using the Harris Interactive data, the number of participants and trips were estimated for the following categories: Bicycle-based recreation: paved roads off-road Camp-based recreation7: recreational vehicles campgrounds/campsite rustic lodges Paddle-based recreation: kayaking (recreational/sea/whitewater) rafting canoeing outdoor swimming (lakes, rivers, natural water bodies) outdoor pool swimming8 Snow-based recreation: cross-country/nordic skiing downhill skiing telemark skiing snowboarding snowshoeing Trail-based recreation: trail running (running on an unpaved trail) day hiking on an unpaved trail backpacking climbing (ice or natural rock) For each state, the weighted Harris data permitted the development of estimates describing the percentage of each state s population having participated in each category in the past year and the average number of trips per person. Estimates were produced for each of the five major categories (trail sports, bicycling, camping, snow sports and paddle sports). Recognizing many people participate in more than one of the five major categories, the total percentage of a population participating in at least one category could not be estimated by summing the results for each sub-category. Doing so would result in significant doublecounting. Overall participation estimates were based on the combined sample of qualified and 7 Camping includes day trips when camping equipment was used. 8 Outdoor swimming and outdoor pool swimming were eliminated during an interim round of analysis based on industry feedback that these activities did not fully fit as an active outdoor recreation, plus their sheer size overshadowed the impact of the other activities. More details are presented in the methodology discussions. 9

non-qualified survey respondents. Recognizing this combined sample, once weighted, reflected the entire state population, and based on a survey question asking if the respondent participated in at least one of the five major categories, overall participation estimates were possible. 9 The participation data may contain an unknown level of double-counting in the number of trips. The number of trips reported for each of the five major categories (bicycle, camping, paddle, snow, and trail-based recreation) is based on the sum of the primary trips reported for each of the sub-categories (for bicycle, sub-categories are off-road and paved-road riding). A bicyclist could have gone on one ride that involved both paved-road and off-road characteristics and counted the ride as two trips one off-road and one paved-road instead of either an offroad or a paved-road bicycle trip. The Harris Interactive survey capped the number of trips that a person could report for all sub-categories at 99. In about 1% of the cases, the number of trips reported for a major category exceeded 99 which indicate some level of double counting across the sub-categories. This is assumed to be based on people who, for example, reported 50 off-road bicycle trips and 50 paved-road bicycle trips which were then summed and reported as 100, thus exceeding the quota of 99. It is possible some respondents reported all of their trips as primary purpose trips, even though they were given the option to designate trips as nonprimary. To help reduce any double counting, the number of trips reported by any one person was capped at 99. It is still possible that even when observations are capped at 99, additional double-counting may exist for trips that summed to a level less than 99. Once the percentages of population who participated were complete, the next step was to estimate the number of participants and trips per state. This was done by multiplying the percentage of population participating in each category by the number of people living in the state under study. Population data were obtained from the U.S. Census Bureau. Total number of trips were then calculated by multiplying the total number of participants by the average number of trips per participant as reported by the Harris data. The estimated number of participants for each state and activity are listed in Table B. Important note: These results are for illustrative purposes only. They describe how the total state active outdoor economic figures were calculated. Do not reprint state activity category level economic figures (for example camping in Nevada) Limitations in sample size create large plausible ranges for some state activity category level economic figures. 9 Question 405 (Appendix II) inquires if respondents participated in any of the five major activities. The subject of the survey was kept generic up to this point to help prevent non-participants from dropping out of the survey. See Q101 for an example of the initial generic approach taken by the survey. 10

Table B. Participation in Outdoor Recreation and Percentage of State Residents Participating in Each Activity Bicycle Camping Fishing* Hunting* Paddling 11 Snow Trail Wildlife Viewing* State Pop. Over 18 years Old State Pop. Over 16 Years Old Alaska 158,526 179,861 176,961 70,776 92,296 92,249 183,504 235,000 477,865 510,956 % of state pop. 33% 35% 35% 14% 19% 19% 38% 46% Arizona 1,151,671 1,067,921 339,417 116,977 320,680 284,229 1,164,256 1,098,000 4,334,537 4,591,010 % of state pop. 27% 25% 7% 3% 7% 7% 27% 24% California 7,467,740 4,862,899 2,212,496 256,563 2,956,918 2,324,238 8,905,167 5,336,000 26,576,445 28,231,375 % of state pop. 28% 18% 8% 1% 11% 9% 34% 19% Colorado 1,212,400 982,690 541,520 156,297 392,451 743,263 1,405,329 1,190,000 3,464,517 3,659,733 % of state pop. 35% 28% 15% 4% 11% 21% 41% 33% Idaho 327,116 460,099 242,717 147,450 161,673 166,909 339,158 379,000 1,027,114 1,089,310 % of state pop. 32% 45% 22% 14% 16% 16% 33% 35% Louisiana 668,978 426,965 728,151 288,019 129,902 76,531 432,671 819,000 3,349,214 3,555,041 % of state pop. 20% 13% 20% 8% 4% 2% 13% 23% Maine 266,782 228,047 205,004 120,909 293,990 240,820 389,424 518,000 1,036,627 1,089,496 % of state pop. 26% 22% 19% 11% 28% 23% 38% 48% Montana 239,448 329,205 205,004 167,110 171,387 133,293 352,206 362,000 717,654 756,877 % of state pop. 33% 46% 27% 22% 24% 19% 49% 48% Nevada 376,009 359,715 115,073 41,286 78,067 104,745 528,203 320,000 1,758,622 1,852,838 % of state pop. 21% 20% 6% 2% 4% 6% 30% 17% New Hampshire 241,100 238,895 142,149 51,116 225,662 254,233 349,570 445,000 1,007,415 1,063,065 % of state pop. 24% 24% 13% 5% 22% 25% 35% 42% New Mexico 322,662 438,031 190,499 103,215 66,906 104,790 584,852 469,000 1,414,523 1,501,226 % of state pop. 23% 31% 13% 7% 5% 7% 41% 31% New York 4,182,579 2,826,543 1,201,981 624,205 1,794,840 1,108,017 3,153,812 3,495,000 14,650,778 15,481,693 % of state pop. 29% 19% 8% 4% 12% 8% 22% 23%

Table B (Continued) North Carolina 1,605,610 1,075,413 803,577 267,376 960,493 349,983 1,704,206 1,868,000 6,550,388 6,933,767 % of state pop. 25% 16% 12% 4% 15% 5% 26% 27% Oregon 773,028 1,016,631 496,071 230,022 368,043 247,491 1,126,627 1,280,000 2,752,582 2,898,000 % of state pop. 28% 37% 17% 8% 13% 9% 41% 44% Tennessee 771,509 945,588 685,603 283,104 362,741 195,498 944,677 1,701,000 4,516,751 4,767,907 % of state pop. 17% 21% 14% 6% 8% 4% 21% 36% Utah 409,425 761,579 375,196 173,991 246,132 310,393 714,023 562,000 1,646,570 1,775,937 % of state pop. 25% 46% 21% 10% 15% 19% 43% 32% Vermont 141,315 102,828 92,832 72,742 64,731 121,803 162,826 282,000 490,745 519,977 % of state pop. 29% 21% 18% 14% 13% 25% 33% 54% Washington 1,522,469 1,685,523 781,336 206,430 555,625 543,671 1,837,100 2,210,000 4,713,756 4,983,250 % of state pop. 32% 36% 16% 4% 12% 12% 39% 44% West Virginia 228,244 402,077 241,750 225,107 114,409 82,620 379,596 500,000 1,427,010 1,495,467 % of state pop. 16% 28% 16% 15% 8% 6% 27% 33% Wisconsin 1,524,056 1,133,121 909,947 578,004 786,937 530,253 1,040,242 2,159,000 4,220,053 4,461,074 % of state pop. 36% 27% 20% 13% 19% 13% 25% 48% Wyoming 111,406 154,731 113,139 63,895 33,061 40,348 139,649 171,000 388,262 409,345 % of state pop. 29% 40% 28% 16% 9% 10% 36% 42% 12

Exclusion of Outdoor Swimming and Pool Swimming After the first draft results were produced, a change was made to the categories under study. Outdoor swimming (lakes, rivers, natural water bodies) and outdoor pool swimming were eliminated based on industry feedback.10 Swimming contributed onequarter of all reported expenditures in the first draft. However, industry reviewers felt it was difficult to classify all swimming categories and dollars as part of active outdoor recreation. Many swimming categories may be part of other categories, such as family vacations, or just a trip to a neighbor s pool primarily to socialize with friends. While these problems may exist to some degree for all trips, and even though questions were added in the survey to separate trips taken for the primary purpose of active outdoor recreation versus secondary-purpose trips, it was regarded as prudent and conservative to exclude swimming from the watersports estimates. The watersports category was redefined as paddle-based recreation. This adjustment is not without concerns. The Harris survey targeted number of surveys (a quota) per state. If a quota was filled with just swimmers, then the other categories are under-represented. Once the swimmers are deleted, the sample size for that state is smaller than the original quota and we do not have as large a representative sample of all the other categories. If swimmers were left off the original questionnaire, the quota would have been filled by participants in the other categories which would have led to better estimates for those categories. Despite this problem, it was felt the results would better reflect the active outdoor recreation and remain more conservative if swimming was left out. Appendix II presents the estimated number of trips and participants for each state, along with other details. Breakouts are provided for sub-categories (such as paddlebased recreation which is divided into kayaking and canoeing). Also presented are estimates on the percentage of the state population that participates in some form of active outdoor recreation at least once annually. Please note that when developing an estimate of the total dollars spent annually for active outdoor recreation, it is not correct to multiply any trips or participation estimate found in this report by the total dollars spent per day or trip. Doing so will overestimate actual expenditures. One needs to consider that not all participants spend money on every trip, and that not all trips are taken for the primary purpose of recreation. Some trips and their expenditures would still occur even if the person could not participate in outdoor recreation. Steps are required to deduct for these events. These adjustments are made later in this report. 3.6 Estimating Expenditures Expenditure profiles were developed based on the Harris Interactive survey data. 11 Average expenditures were developed from the qualified set of survey respondents as 10 Outdoor swimming and outdoor pool swimming were eliminated during an interim round of analysis based on industry feedback that these activities did not fully fit as an active outdoor outdoor recreation. More details are presented in the methodology discussions. 11 Questions 516 to Q560 (equipment and durables), and Q631 thru Q690 in Appendix IV. 13

explained earlier. Qualified respondents are those who participated in one of the recreation categories tracked in this report and spent money on outdoor recreation. The equipment and durable expenditure estimates developed from Harris Interactive outdoor data estimate the mean amount spent per year per participant. This was done for each expenditure item queried in the survey. Data were collected for expenditures made by the respondent for his or her personal use and for items purchased for minors under 18. Minors were excluded from the survey and querying adults about expenditures for children helped capture some of the expenditures made for minors. In some cases, the sample size for specific items at the state level were low, which could result in estimates relatively far from the true mean. To help reduce such variance in the results, an adjustment was made. Whenever the state estimate for a specific expenditure item was 25 percent greater than the national estimate for the same item, the state estimate was replaced with the national estimate. When state expenditure estimates were 25 percent lower than the national estimates, they were left as-is to help maintain a level of conservatism. Appendix II presents, per state, the typical expenditure per trip for trip-related expenses plus the average annual expenditures for equipment and services on a perperson basis. 12 3.6.1. Equipment and Services Expenditures Total expenditures were calculated by multiplying the average expenditure per person by the total number of participants. For annual equipment & services expenditures, an extra step was added to adjust them downward to only reflect trip expenditures made for the primary purpose of active outdoor recreation. This was based on questions that inquired approximately what percentage of your total (spent) do you think you would have spent even if you had not participated in the category in question (see Q555, Appendix V, for an example). The reported percentage was deducted from the total reported expenditure. An additional step was also made that adjusted the equipment and services expenditures downward to only reflect the percentage of survey respondents who reported making an expenditure for the category in question (see Q520, Appendix V, for an example). This study developed economic impact estimates for the major forms of active outdoor recreation of interest to the OIA and OIF. By definition, these exclude categories that rely on the primary use of internal-combustion engines such as motorsports (off-road), boating, snowmobiling, etc. While the purchases of vehicles and other high-end purchases were discounted in this study by treating them as outliers, trip costs associated with vehicles, such as fuel, were not excluded. Treatment of Durable Equipment Outliers: The Harris survey did not specifically inquire about respondents purchases of recreational vehicles and other big ticket purchases. To exclude big ticket purchases that could be used for other categories not related to active outdoor recreation, observations reporting equipment expenditures 12 These are based on qualified respondents, which are people who took at least one trip and reported expenditures as part of the trip. 14

exceeding $10,000 were capped at $10,000. Expenditures for apparel, accessories and services were capped at $2,000. The only significant outliers in the expenditure data, however, were in the camping equipment category and were believed to have been made for recreational vehicles and campers. Recognizing these vehicles can be used for many activities besides active outdoor recreation, industry reviewers recommended that observations exceeding $2,000 in the camping equipment category be capped at $2,000. This adjustment was made to the data to help ensure the results remain a conservative estimate of the impacts from active outdoor recreation only and minimize the potential inclusion of dollars that can also be attributable to other activities. 3.6.2. Trip-Related Expenditures For trip related expenditures, profiles of the average amount spent per trip for each state were developed based on the survey responses. These profiles provided expenditure estimates for each of the five major categories. Table C presents a sample of expenditures developed from the survey. The detailed results for day trips and overnights for each state can be found in Appendix II. Table C: Sample of Trip Expenditures Profile: Average Day Trip Expenditures, Per Trip, National Trail Bicycle Camp Snow Paddle Mean N Mean N Mean N Mean N Mean N Food & Drink $12 1622 $8 1356 $63 1011 $32 727 $18 791 Transportation $13 1622 $7 1356 $60 1011 $33 727 $17 791 Recreation, Entertainment & Activities $6 1622 $4 1356 $30 1011 $36 727 $20 791 Souvenirs, Gifts & Other Miscellaneous $5 1622 $2 1356 $16 1011 $9 727 $5 791 Trip expenses were calculated by multiplying the total number of trips reported for a state by the state s respective expenditure profile. Using data from survey questions inquiring about the percentage of all trips taken for the primary purpose of recreation, all expenditures made for primary trips were included. Typically, 70 percent to 95 percent of all trips were reported as primary-purpose trips. Based on input from industry reviewers, 10 percent of non-primary trips were added back in. This was allowed as some level of secondary trips would not be taken if active outdoor recreation was not available. The availability of active outdoor recreation is reasonably expected to raise the marginal benefits for many trips up to a point that, without the recreational opportunity available, the trip would otherwise not have been taken. In addition, the presence of the recreational category as part of another trip is expected to increase the cost of the trip and the additional amount would be fully attributable to outdoor recreation. Without data identifying the dollars spent on secondary trips attributable to recreation, 10 percent is considered a reasonable amount to allocate. 15

Just as done for expenditures for durable items, a step was taken to adjust trip expenditures downward to only reflect the percentage of survey respondents who reported making an expenditure for the category in question (see Q665, Appendix V, for an example). Within a state, expenses made by residents of other states were included as well (see Q690, Appendix V, for an example). For visitors from outside the state, only their expenditures made out of their states of residence were included. This adjustment was made based on data from Q690. An adjustment was also made to only include expenditures for people who reported making purchases for a particular item. Depending on the category, 20 percent to 90 percent of the participants reported making expenditures as part of their trip. Trail sports and bicycling typically had the lowest rates of expenditures among participants. If, for a given state, only 51 percent of the respondents reported spending money on their categories, then only 51 percent of the reported trips were used in the expenditure equations. Whenever the state estimate for a specific trip expenditure item was 25 percent greater than the national estimate for the same item, the state estimate was replaced with the national estimate. When state expenditure estimates were 25 percent lower than the national estimates, they were left as-is to help maintain conservative results. This same adjustment was made for equipment/durable goods as explained earlier. Treatment of Trip-Related Outliers: Trip expenditures that exceeded $500 for a particular category (food, transportation, etc.) for day trips were capped at $500, and overnight trip expenditures were capped at $2,000 per category. This results in a maximum of $2,000 per day trip and $10,000 for overnight trips. There were very few observations that required this adjustment. See Appendix II for detailed total and per-trip expenditures per state. 16

Assumptions and Caveats to the Expenditure Data: 1) Thirty-five percent of the survey respondents who reported overnight camping indicated they typically do not take part in one of the other four major categories while camping (trail sports, bicycling, snow sports or water/paddle sports). While some of them may participate in wildlife-based recreation (hunting, fishing and wildlife viewing), many of them may not. It should be noted that a small percent of campers may not be regarded as participating in a form of active outdoor recreation and may just relax at the campsite. 2) Survey respondents for the expenditure questions were limited to qualified respondents defined as persons who participated in recreation during the previous 12 months and who indicated they spent some amount of money on recreation during the previous 12 months. Thus, the resulting participant expenditure sample is not representative of all recreation participants since those who spent no money on recreation during the previous 12 months were not included. For estimation of total expenditures, mean expenditures were aggregated only over the total number of estimated participants who spent some amount of money on recreation during the previous 12 months. 3) The survey was limited to people 18 years old and older. Expenditures made by minors are excluded, unless the purchase was made by an adult for a minor. Recognizing the high level of disposable income among youth, an unknown level of expenditures are left out of this study. 4) As with any survey, there is some level of natural error associated with not finding people who participated in an activity or purchased specific items. This is common for activities and items of lesser popularity. Examples include non-resident trips to Alaska for snow sports and people buying rarer items such as souvenirs as part of cross-country ski trips. Larger sample sizes tend to diminish this problem, but even with major surveys such as the multi-million U.S. Fish and Wildlife Service s 2001 National Survey on Fishing, Hunting and Wildlife-Associated Recreation, this problem still occurs. This problem often results in low numbers of estimated trips, lower than what may really be occurring, which leads to underestimated expenditures and economic impacts. Appendix IV presents the errors associated with the trip estimates which can be used to identify those estimates with a smaller or greater range of error. As noted elsewhere in this document, when the average expenditure for a state, for a specific item, was 25%, or more, greater than the national average, it was replaced by the national average to help reduce the chance of overestimating impacts. When the average was 25%, or more, less than the national average, it was left as-is to maintain this report s conservative nature. Table D present the expenditures per state for all five activities. Appendix II presents greater trip, participation and expenditure details. Important note: These results are for illustrative purposes only. They describe how the total state active outdoor economic figures were calculated. Do 17

not reprint state activity category level economic figures (for example camping in Nevada). Limitations in sample size create large plausible ranges for some state activity category level economic figures. 18

Table D. Total Expenditures Per State Trail Bicycle Camping Snow Paddle TOTAL: Alaska All Expenditures: $108,302,703 $47,124,200 $358,816,872 $153,080,451 $48,164,128 $715,488,353 Travel-related: $98,752,091 $30,484,095 $321,790,988 $135,458,392 $38,130,948 $624,616,515 Equip. & Accessories: $9,550,611 $16,640,105 $37,025,884 $17,622,059 $10,033,180 $90,871,839 Arizona All Expenditures: $315,205,392 $917,668,089 $2,323,206,783 $441,746,986 $100,877,944 $4,098,705,194 Travel-related: $276,717,954 $796,672,214 $2,117,774,530 $372,603,733 $78,172,982 $3,641,941,414 Equip. & Accessories: $38,487,438 $120,995,875 $205,432,253 $69,143,252 $22,704,962 $456,763,780 California All Expenditures: $3,270,101,916 $6,472,609,248 $7,923,526,625 $3,415,603,752 $3,512,972,401 $24,594,813,942 Travel-related: $2,831,088,350 $5,391,652,106 $7,138,675,299 $2,933,967,492 $3,155,868,590 $21,451,251,837 Equip. & Accessories: $439,013,566 $1,080,957,142 $784,851,327 $481,636,260 $357,103,811 $3,143,562,105 Colorado All Expenditures: $974,466,454 $858,474,741 $1,936,230,560 $2,314,753,272 $408,464,332 $6,492,389,359 Travel-related: $891,219,604 $727,642,554 $1,807,600,172 $2,080,869,863 $349,743,329 $5,857,075,523 Equip. & Accessories: $83,246,850 $130,832,187 $128,630,388 $233,883,409 $58,721,003 $635,313,837 Idaho All Expenditures: $266,201,300 $52,578,051 $1,056,015,287 $305,475,711 $50,898,900 $1,731,169,248 Travel-related: $257,685,368 $27,314,174 $997,594,958 $282,592,044 $38,756,363 $1,603,942,906 Equip. & Accessories: $8,515,932 $25,263,877 $58,420,329 $22,883,667 $12,142,537 $127,226,342 Louisiana All Expenditures: $439,363,355 $676,843,771 $990,429,492 $36,963,786 $110,820,770 $2,254,421,174 Travel-related: $406,860,532 $588,300,131 $916,854,796 $11,927,426 $93,511,758 $2,017,454,644 Equip. & Accessories: $32,502,823 $88,543,640 $73,574,696 $25,036,360 $17,309,011 $236,966,530 Maine All Expenditures: $359,762,341 $474,351,661 $934,948,089 $490,918,556 $163,943,225 $2,423,923,871 Travel-related: $341,225,456 $447,034,173 $893,415,651 $468,250,037 $147,244,126 $2,297,169,443 Equip. & Accessories: $18,536,885 $27,317,487 $41,532,438 $22,668,519 $16,699,099 $126,754,428 Montana All Expenditures: $313,134,500 $58,730,785 $673,800,845 $226,760,462 $81,636,806 $1,354,063,398 Travel-related: $306,343,607 $42,788,016 $643,129,761 $192,487,252 $78,763,076 $1,263,511,713 Equip. & Accessories: $6,790,893 $15,942,769 $30,671,084 $34,273,209 $2,873,730 $90,551,685 Nevada All Expenditures: $483,474,548 $152,157,027 $474,997,133 $323,169,023 $31,997,769 $1,465,795,500 Travel-related: $457,478,651 $95,297,525 $409,640,417 $306,438,009 $31,723,727 $1,300,578,329 Equip. & Accessories: $25,995,897 $56,859,502 $65,356,716 $16,731,014 $274,042 $165,217,171 New Hampshire All Expenditures: $453,736,977 $77,474,447 $2,426,550,441 $544,989,692 $78,247,620 $3,580,999,177 Travel-related: $428,419,098 $52,782,243 $2,391,380,056 $504,216,727 $68,094,536 $3,444,892,659 Equip. & Accessories: $25,317,879 $24,692,204 $35,170,385 $40,772,965 $10,153,085 $136,106,518 New Mexico All Expenditures: $634,222,240 $308,565,537 $769,602,333 $420,117,086 $40,773,172 $2,173,280,369 Travel-related: $582,106,997 $273,995,523 $719,069,324 $407,936,219 $36,728,069 $2,019,836,133 Equip. & Accessories: $52,115,243 $34,570,014 $50,533,010 $12,180,867 $4,045,103 $153,444,236 New York All Expenditures: $1,530,857,961 $1,143,671,530 $4,062,741,906 $1,673,914,228 $1,010,244,766 $9,421,430,390 Travel-related: $1,314,582,032 $803,455,156 $3,552,289,125 $1,367,891,291 $860,172,581 $7,898,390,186 Equip. & Accessories: $216,275,929 $340,216,374 $510,452,780 $306,022,936 $150,072,185 $1,523,040,204 19

Table D. Total Expenditures Per State (continued) N. Carolina All Expenditures: $983,654,662 $992,975,959 $1,866,131,324 $559,762,626 $363,107,776 $4,765,632,346 Travel-related: $906,742,482 $812,291,466 $1,722,037,751 $506,997,677 $309,440,907 $4,257,510,283 Equip. & Accessories: $76,912,181 $180,684,492 $144,093,573 $52,764,949 $53,666,869 $508,122,064 Oregon All Expenditures: $589,736,126 $464,435,058 $1,804,331,362 $515,892,191 $114,200,258 $3,488,594,995 Travel-related: $535,285,480 $394,499,737 $1,685,927,670 $504,671,122 $94,637,311 $3,215,021,320 Equip. & Accessories: $54,450,646 $69,935,321 $118,403,692 $11,221,069 $19,562,947 $273,573,676 Tennessee All Expenditures: $502,199,019 $1,191,033,224 $1,648,524,675 $51,322,958 $173,795,358 $3,566,875,234 Travel-related: $473,810,704 $1,062,463,171 $1,566,136,682 $23,070,919 $141,111,305 $3,266,592,781 Equip. & Accessories: $28,388,315 $128,570,053 $82,387,993 $28,252,039 $32,684,053 $300,282,453 Utah All Expenditures: $880,710,914 $630,930,354 $1,071,081,986 $463,940,666 $99,675,087 $3,146,339,006 Travel-related: $857,742,165 $581,937,350 $1,005,614,970 $440,415,672 $94,666,978 $2,980,377,134 Equip. & Accessories: $22,968,749 $48,993,005 $65,467,015 $23,524,994 $5,008,109 $165,961,872 Vermont All Expenditures: $292,067,350 $71,102,769 $861,608,608 $1,019,905,343 $50,915,952 $2,295,600,022 Travel-related: $283,743,048 $61,136,130 $844,634,980 $1,006,550,514 $45,788,841 $2,241,853,512 Equip. & Accessories: $8,324,302 $9,966,639 $16,973,629 $13,354,829 $5,127,112 $53,746,510 Washington All Expenditures: $1,009,450,111 $1,742,359,916 $3,582,187,506 $696,716,166 $153,994,573 $7,184,708,271 Travel-related: $861,462,217 $1,565,017,162 $3,318,209,374 $586,227,652 $88,175,746 $6,419,092,151 Equip. & Accessories: $147,987,895 $177,342,754 $263,978,132 $110,488,513 $65,818,827 $765,616,121 West Virginia All Expenditures: $278,782,192 $345,107,278 $2,663,589,898 $540,443,677 $126,623,885 $3,954,546,931 Travel-related: $265,217,153 $329,150,226 $2,590,342,404 $519,576,348 $121,047,181 $3,825,333,311 Equip. & Accessories: $13,565,039 $15,957,052 $73,247,494 $20,867,330 $5,576,704 $129,213,619 Wisconsin All Expenditures: $298,109,229 $541,199,770 $3,962,559,272 $305,810,453 $406,211,021 $5,513,889,744 Travel-related: $271,061,846 $469,803,157 $3,766,593,093 $282,480,138 $345,571,095 $5,135,509,329 Equip. & Accessories: $27,047,383 $71,396,612 $195,966,179 $23,330,314 $60,639,926 $378,380,415 Wyoming All Expenditures: $347,704,643 $102,208,625 $2,565,621,595 $182,390,235 $37,903,958 $3,235,829,057 Travel-related: $344,190,598 $96,849,633 $2,539,814,155 $173,202,546 $35,137,010 $3,189,193,941 Equip. & Accessories: $3,514,046 $5,358,993 $25,807,440 $9,187,689 $2,766,948 $46,635,116 20