May 15, 1997 AES Project Outline

PROJECT TITLE: Economic Impacts of Recreation and Tourism

PRINCIPAL INVESTIGATOR :

Daniel J. Stynes, Professor, Dept. of Park,Recreation and Tourism Resources

COOPERATORS

Dennis B. Propst, Assoc. Professor, Dept. of Park,Recreation and Tourism Resources

Edward Mahoney, Assoc. Professor, Dept. of Park,Recreation and Tourism Resources

Travel Tourism and Recreation Resource Center, Michigan State University

START DATE: July 1, 1997Probable Duration: 5 years

PROBLEM/IMPORTANCE: Recreation and tourism are important elements of Michigan’s economy. While there are statewide estimates of some $8 billion in tourist spending and numerous local or activity-specific studies estimating economic impacts of particular activities, parks, festivals or events, we still lack reliable estimates of the importance of particular tourist market segments to subregions of the state. Statewide tourism estimates have been too aggregate to shed much light on particular markets or subregions, while local and site specific studies have been both too incomplete and too variable in methods and accuracy to assemble a more complete and consistent picture. Improvements in economic databases and analysis tools over the past five years now permit a more consistent and standardized approach for estimating the economic significance and impact of recreation and tourism in Michigan. Using satellite accounting methods, recreation and tourism surveys, regional economic models and government economic databases, the proposed study will compile comprehensive estimates of recreation and tourism’s economic significance to the state and local areas. The study will also identify the relative contribution of key market segments to particular subregions of the state.

Economic impacts of various activities and policies continue to carry a great deal of weight in both public and private decisions. To be credible, estimates of the economic impacts of recreation and tourism must be based on sound data and methods and must be firmly grounded in the national system of economic accounts. To be more useful for policy, marketing and management decisions, economic estimates must be more disaggregated to identify the specific impacts of particular actions or market segments on particular regions of the state. Economic estimates for different regions, activities and market segments must be comparable if they are to be used to make resource allocation decisions.

The proposed project will pursue four broad areas of research to further our understanding of the regional economic impacts of recreation and tourism. First, we will apply the existing tools to a variety of types of tourism and regions to estimate economic impacts of various recreation and tourism activities throughout Michigan. These applied studies will provide valuable information to planners and decision makers throughout the state, while also contributing to a more general understanding of the regional economic consequences of tourism. A byproduct of these studies will be firmer estimates of the relative importance of different types of tourism and market segments to the economy of the state and its subregions. Studies are already underway to estimate economic impacts of boating, seasonal homes, and state forest campgrounds. These applied studies will be continued, generally with funding from external sources. We will seek partners to support economic studies for key segments that have not been covered well to date, e.g. beach users, downhill skiing, snowmobiling, golf, festivals and events.

Under a second key objective, estimates of tourism activity and spending from these studies and other secondary sources will be organized into a comprehensive and consistent set of accounts that identifies the relative contribution of particular tourist segments to overall recreation/tourism activity (person days) and spending as well as the overall contribution of tourism to gross state product. Knowing the relative importance of camping, boating, downhill skiing, family vacation travel, seasonal homes, hunting, and fishing to the state and individual subregions is critical to allocations of resources and marketing efforts. Tracking trends in tourism similarly requires a better understanding of which markets are growing or declining and why. A consistent accounting framework is essential to evaluate the validity of tourism impact estimates and to compare tourism impacts with those of other sectors. Tourism spending may be organized into the national system of accounts by employing satellite accounting procedures developed recently by the WEFA Group and the World Travel and Tourism Council (WTTC 1993). Hawaii was the first state to adopt this approach (WTTC 1996).

The third area of further research is to examine the regional economic structure of various subregions of the state. The initial work in this area is being carried out for the Eastern Upper Peninsula in a SAPMINR project. Over the next five years, we would complete similar assessments for other regions of the state. Using secondary data, we also will examine the economies of a number of unique tourism regions around the country (e.g. Hawaii, Las Vegas, Florida Keys, Branson Missouri, Colorado ski country). The purpose of these studies will be to compare the economic structure of regions with varying degrees of tourism dependency. We will also examine differences in the economic consequences of distinct types of tourism, e.g. highly seasonal vacation travel, short-term festivals and events, seasonal homes, and outdoor recreation.

Finally, we will continue to investigate a number of technical and methodological issues involved in assessing the economic impacts of tourism. There has been little research to examine whether various errors in economic databases or model assumptions result in significant biases in tourism impact estimates. There have also been no attempts we know of beyond those in the MI-REC system (Stynes 1996d) to refine or adapt economic impact tools for the special circumstances posed by recreation and tourism, for example, seasonality, part time employment, and spending predominantly in retail trade and service sectors. One area deserving further attention is the validity and accuracy of estimates of the induced effects of tourism spending. Our studies have found that the vast majority of the multiplier effects from tourism are induced rather than indirect. Many have questioned both the methods for estimating induced effects and their appropriateness for particular policy decisions.

Tourism organizations and regional planning and economic development groups continue to demand better information about tourism’s economic impacts. Decisionmakers are increasingly seeking reliable estimates that are backed by sound data and models. Decisions and policies generally require specific information about the economic impacts of specific actions or market segments on particular subregions of the state.

OBJECTIVES

  1. Estimate the economic impacts of key recreation and tourism activities and market segments on subregions of Michigan.
  2. Assemble comprehensive estimates of tourism’s contribution to the Michigan economy including the relative importance of key activities, regions, and market segments.
  3. Describe the regional economic characteristics of different types of tourism-dependent regions, contrasting these economies with those of regions with alternative economic bases.
  4. Evaluate selected assumptions of regional economic models and their applicability to recreation and tourism assessments. Refine the tools to adjust for invalid assumptions and common shortcomings of the underlying economic databases.

RESUME OF PREVIOUS INVESTIGATIONS

Over the past ten years there have been numerous studies of the economic impacts of recreation and tourism. Recent published bibliographies (Fredick, n.d.) and others posted on the WWW (e.g. at NRPA and M.I.G. sites) include a couple hundred listings in the past ten years and these lists do not include large numbers of unpublished reports, which constitute the majority of studies. These applied studies fall into four categories :

(1)impacts of existing state and national parks, forests or water-based facilities (e.g. Bergstrom et. al 1990, Jackson et. al. 1994, the National Park Service’s Money Generation model, 1990),

(2) impacts of particular recreation and tourism market segments such as boating, hunting and fishing (e.g. Stynes et. al. 1997, USDI, 1991 National Survey of Hunting and Fishing),.

(3) impacts of tourism on national, state and regional economies (Mak 1989, Frechtling 1994, Michigan Jobs Commission 1996, Stynes 1996a) , and

(4) impacts of festivals and special events (Getz 1991, 1994).

This applied research tends to be one shot case studies with limited attention to generalizable methods or results. Some of these are spending studies, others examine secondary effects using “off-the-shelf” multipliers (U.S. Dept. of Commerce, 1992) or input-output models. Studies vary widely in quality, methods and coverage (Fleming and Toepper 1990). Most focus on visitor trip spending although some have addressed the impacts of durable goods purchases (Propst et. al. 1992) and government operations (Stynes and Rutz 1995).

A search of the CRIS project database uncovered about 70 active or recently expired projects dealing with some aspect of the economic impacts of recreation or tourism. A handful of these are directly related to some of our planned work and we are contacting those investigators for further information and copies of reports that we do not already have. In particular, Uysal has developed procedures for studying festivals and events in Virgina, Nickerson has completed statewide tourism impact estimates for Montana, and Strauss has developed spending profiles for several types of tourists in Pennsylvania.

Recreation and tourism analysts have generally accepted standard regional economic analysis methods (Richardson 1972, Mierynk 1965) and increasingly are making use of input-output models. Modeling systems like IMPLAN and MI-REC are better known and more widely used in outdoor recreation than tourism. Aside from problems in properly applying and interpreting regional economic concepts and tools, the greatest obstacle to sound estimates of the economic impacts of recreation and tourism is the lack of good data on the levels of recreation and tourism activity.

Generally, recreation and tourism activity and spending must be estimated in visitor surveys (Sheldon 1990). Inconsistency in survey methods and small samples in many cases make it difficult to compare spending estimates for particular segments across studies (Spotts and Mahoney 1991). Survey estimates of spending are likely to be upward biased due to telescoping problems, non-response bias, and sampling problems. Recreation and tourism are very diverse activities with high variances in spending across regions and market segments.

Government economic data series (e.g. lodging room use taxes, and sales, income and employment by sector) offer considerable potential to supplement visitor surveys and help validate economic impact estimates. Stynes and Rutz (1995) have found that lodging receipts can provide a useful check on some of the biases in surveys and park use data. Sales and tax data can be used to ground tourism economic impact estimates within existing systems of economic accounts.

One of the more important recent developments in assessing tourism’s economic importance is the satellite accounting methodology developed by the WTTC (1993) in collaboration with the WEFA group. Their satellite accounting system captures tourism economic activity within the national system of accounts. These procedures offer a promising approach of combining survey and secondary data to produce more reliable and complete estimates of tourism’s economic impacts, while also embedding the estimates within a standard and widely accepted system of accounts. The WTTC system focuses on the direct effects of tourism, including consumer spending as well as government spending and capital expenditures. The WTTC system does not cover multiplier effects. WTTC estimates that tourism constitutes about 10.2% of gross domestic product in the U.S. and 24.3% of gross state product in Hawaii.

Multipliers and multiplier effects are also important in understanding the linkages between tourism and various sectors of the regional economy. Regional economic impact modeling tools have received limited but growing use in recreation and tourism (Frechtling 1994), where economic data, regional economic multipliers, and models have been frequently misused and misunderstood (Archer 1984, Crompton 1993 ). Stynes and Propst’s (1996) MI-REC system makes it relatively easy to estimate economic impacts when the number of visitors and their spending patterns are known, that is, if one is willing to accept the standard assumptions of input-output models and the underlying economic databases. While there are a number of cases where the assumptions do not completely hold, violations of the assumptions do not appear to significantly influence aggregate results. When analysts are interested in more specific estimates for particular sectors, errors can however be substantial.

The proposed study will build upon work carried out in three recently expired projects. We will make extensive use of the MI-REC economic impact estimation tools developed by Stynes and Propst, utilizing IMPLAN and recent databases for Michigan to study the economic structure of regions in Michigan. The MI-REC recreation and tourism spending profiles assembled to date will be extended to other activities and market segments such as winter sports, seasonal home trips, and visits to festivals and events. These spending databases will be used to develop economic impact estimates for these activities. Consumer price indices will be employed to adjust various spending datasets to a common year.

The study will also build on both recreation activity estimation and tourism spending models developed by Stynes (1996) for the Lake States Forest Assessment. Five lodging-based segments were used in that study as a basis for estimating recreation activity and tourist spending by county for all Michigan counties circa 1990. Visitors staying in motels, campgrounds, seasonal homes, or with friends and relatives, and day users from outside the county comprise the five tourist segments used to predict spending. A lodging inventory, spending profiles for each segment, lodging room use taxes and a set of recreation activity and occupancy rate parameters estimated from recent surveys complete the data requirements for the models. This study will update the activity and spending estimates using 1995 lodging tax data and more recent estimates of various model parameters.

The Michigan Jobs Commission (1997) estimates tourist spending in Michigan at around $8 billion for 1994, although their methods likely overstate some types of tourism while not fully representing others (e.g. seasonal home and outdoor recreation trips). Their estimates only cover spending of tourists on trips of 50 miles or more or overnight and therefore do not include local recreation spending, durable goods purchases or tourism-related construction, infrastructure and other capital expenditures. The U.S. Travel Data Center (1992) estimated tourist spending in Michigan at $6.7 billion in 1990. Stynes (1996) derived roughly similar statewide estimates for 1990. However, using a more disaggregate approach, he uncovered wide differences in estimates for particular counties and market segments. Industry and USTDC estimates tend to be skewed toward urban areas and air travel related spending, while important rural tourism segments such as seasonal homes, camping and outdoor recreation are not covered very well. Rural economies in Michigan are much more dependent on tourism than urban areas, and they also depend on significantly different types of tourism.

EXPERIMENTAL PROTOCOL

As this is a broad umbella project, the methods can only be presented in terms of general approaches to each objective. The specific protocols will depend on the particular studies undertaken, much of which will depend on external funding. Protocols for several of the initial studies are included in separate proposals (e.g. three SAPMINR proposals). Specific models for estimating recreation activity and tourist spending by county are documented in Stynes (1996a and 1996b).

Objective 1: Estimates of economic impacts of particular recreation and tourism activities or markets involve three primary steps, as detailed in the MI-REC system manual (Stynes and Propst 1996). One must first define the action to be evaluated and estimate the changes in the number and types of tourists resulting from the action. In many applications, this is the most difficult step as most regions and tourism organizations lack reliable estimates of the number of tourists. While this project will not attempt to develop demand models, we will develop and test procedures for estimating the numbers and types of visitors in selected studies. One of the more challenging areas will be estimating visitation to festivals and events (Getz, 1991, 1994), starting with the Venetian Festival in St. Joseph Michigan in June, 1997.

The second step is to estimate direct spending by these tourists. Following the MI-REC system, we will estimate spending profiles on a party-day or trip basis by market segment and multiply spending averages by number of visitors in each segment to estimate total spending. Spending profiles are developed with an eye to generalizing these profiles across applications. The project will undertake some primar spending studies (generally in cooperation with other studies) and develop profiles for various tourism segments from secondary sources. Efforts to model spending and explain variations across different locations will also continue, generally using standard linear and non-linear regression methods. Spending profiles from primary data, secondary data, and modeling will be added to the MI-REC system for use in future applications.