Visitor spending can be important to county economies, but it is difficult to estimate visitors’ expenditures on a county level. The Institute for Tourism and Recreation Research (ITRR) has conducted several statewide nonresident travel studies, including a winter study in 1998, a summer study in 1996, and a yearlong study in 1993. Although the studies were done at the state level, they do contain data that can be used to try to estimate nonresident expenditures in various counties. This report provides a description of how those estimates can be made and identifies limitations and gaps in the data currently available from ITRR and other sources. It uses Madison and Lincoln Counties for illustrative purposes.

Visitors to Counties

Visitors in this case are defined as people traveling for a variety of reasons, including vacation, visiting friends and relatives, and business. ITRR does not survey people traveling in commercial vehicles, busses, trains, or tour groups. In order to understand the role visitors play in an economy, it is useful to look at the amount of money they spend in that economy. The definition of a visitor is usually determined by the geographic area of the economy that one is analyzing. In the case of estimates of the economic impact of visitors for the state of Montana, ITRR focuses on visitors from outside the state, because these people bring "new money" into the state’s economy.

Ideally, at the county level, an analysis of visitors’ spending in the economy would include Montana residents from other counties as well as visitors from outside the state. Unfortunately, resident expenditure data are not available. Therefore, at the current time, county-level expenditure estimates can only be provided for visitors who are not residents of Montana. ITRR is in the midst of a yearlong resident travel survey, which might provide additional information regarding the impact of spending by out-of-county residents on various counties.

Data

Secondary Data Availability

Published economic data

It is fairly straightforward to estimate the "size" of most industries, like mining or wood products manufacturing. There are published reports of income and employment by industry and reports on industries’ sales. Unfortunately "travel" is not an industry. Instead, it is an activity that crosses the boundaries of many industries, such as hotels and lodging, restaurants, service stations, grocery stores, amusement services, and various retail stores. Each of the industries that are impacted by travel are also supported by other types of activities—for example, you or I buying our weekly groceries or filling up the car to go to work also has an impact on the industries affected by travel. So, estimating the magnitude of travel activity cannot be done easily through published economic data.

County-level economic analyses

There are no available comprehensive estimates of travel expenditures for Madison and Lincoln Counties, nor are we aware of such estimates for other counties in Montana. There are several reports that provide information about the economics impact of various activities, such as hunting or fishing. But, these do not provide a complete picture of all travel activity taking place in an area.

Visitation estimates

There are a few estimates of visitation figures that can be obtained from land management agencies, such as the Forest Service, but it is not always possible to break these estimates down to the county levels and these estimates do not differentiate among residents (county or state) and nonresidents. Also, because of the dispersed nature of many recreational activities, it is difficult to estimate the number of visitors in an area. The only consistently reliable estimates appear to be those for developed campground sites overseen by campground hosts. There are a few instances where agencies have made further efforts to estimate use in a particular area or by a particular group of users (for example, some have done snowmobile counts in certain areas). In sum, data available from various agencies does not provide enough information to estimate total visitation in a county.

Accommodations tax

The accommodations tax (4 percent of hotel and non-federal campground sales) collected by the state of Montana does provide some indication of activity within the "travel" industry. It is somewhat useful at the county level, where we can be reasonably assured that the vast majority of activity is attributable to people from outside the county. It is less useful at the state level because state residents rent rooms away from their homes. In the least, the accommodations tax tells us the size of the lodging industry in terms of sales.

We explored the feasibility of using the accommodations tax to estimate total expenditures in a county, focusing specifically on Lincoln and Madison Counties. These counties each collected approximately $73,000 in accommodations taxes in 1997, indicating that sales in the hotel and campground industries were approximately $1.8 million. The amount each collected was about 0.8% of the total accommodations tax collected in the state. One option in arriving at an estimate of nonresident spending in a county is to apply this percentage to total statewide nonresident expenditures. For example, in 1997, approximately $1.5 billion was spent by nonresidents in Montana. Applying 0.8% to this figure would mean about $11.5 million was spent in each Lincoln and Madison Counties. There are several limitations to using this approach:

• First, we do not know if and how the hotel and camping industries in these counties might differ from other areas in the state. The average room rates in these counties might be higher or lower than other areas, or there might be a different mix of camping and hotel nights. Each of these variables would affect the number of visitors represented by the accommodations tax and the amount spent by those visitors.
• Second, both residents and nonresidents pay the accommodations tax. Therefore, by using the tax as an indicator, one assumes that there is an equal distribution of resident and nonresident accommodations expenditures across the state. That assumption may or may not be valid, and there are no data to verify it.
• Third, using the tax to distribute expenditures among counties assumes that visitors spend similar amounts in each county. But, a portion of the $1.5 billion spent by nonresidents throughout the state was spent on goods and services not available in every county. Some counties have many more facilities, stores, and services where visitors can spend money. Therefore, we would not expect spending associated with a certain amount of lodging expenditures in one area to equal spending associated with the same amount of lodging expenditures in another area. Even though a person stays in a county and has a hotel expenditure, that person might not spend additional money in that place similar to how money is spent elsewhere in the state. For example, we would expect people to spend more in a place such as Flathead County, which has a ski resort, lake resorts, and many restaurants and shops, than a county such as Phillips County, where there are less opportunities for visitors to spend money.
• Finally, use of the accommodations tax to allocate spending to counties ignores the expenditures made by those staying with friends and relatives, those staying at federal campgrounds, and day visitors. These groups account for a substantial portion of nonresident spending in the state and are likely to have different spending patterns than others (see McMahon and Aldred Cheek, 1998 for more information on spending patterns of nonresident visitors).

The accommodations tax provides a gross comparative estimate of the amount of activity in various counties in the state. For example, it is instructive to know that Lincoln and Madison Counties each collect about 0.8% of the bed tax, while other counties collect up to 22% (Gallatin County). However, its usefulness beyond this rough comparison is very limited.

Using ITRR Nonresident Estimates at the County Level

Two methods

There are two pieces of information within ITRR nonresident studies indicating that visitors stopped and spent money in a particular place. First, nonresident visitors reported where they spent nights on their trip (up to six nights reported in the summer and shoulder seasons and up to 12 nights reported in the winter season). Second, respondents reported the location of each of their expenditures. These two pieces of information were used to estimate the percent of nights and the percent of dollars spent in each of the counties as compared to total nights and total dollars spent statewide. These percentages were then applied to statewide totals to arrive at estimates for the counties. Figures for Lincoln and Madison Counties are provided in Tables 1 and 2. The number of observations associated with each percentage is also provided to demonstrate the availability of data to estimate values for these two counties.

Table 1: Expenditures in counties estimated by nights spent in counties (ITRR surveys)

Table 2: Total expenditures in counties estimated by percent spending in counties (ITRR surveys)

Survey data also provide information about the types of expenditures that visitors make. Table 3 provides a percentage breakdown by spending category for Lincoln and Madison Counties. Note that the figures from the winter and shoulder season surveys that were used to help determine the estimates in Table 3 are based on very few observations.

Table 3: Visitor spending patterns in Lincoln and Madison Counties

Analysis of Using ITRR Data at County Level

Based on ITRR survey data, estimated nonresident expenditures are $9.0 to $18.6 million (0.6% to 1.2% of the state total) in Madison County and $14.7 million to $26.0 million (1.0% to 1.7% of the state total) in Lincoln County. The differences seen between Tables 1 and 2 highlight the difficulty associated with accurately estimating expenditures at the county level based on statewide data. The following paragraphs provide some explanations for these differences and suggest that the estimates based on expenditures in the counties (rather than nights spent) are more accurate.

Expenditures in counties estimated by nights spent in counties (Table 1)

Using percent of nights as an indicator, like using the accommodations tax, assumes that visitors spend similar amounts in each county. But, we would not expect spending associated with a certain number of nights spent in one place to equal spending associated with the same number of nights in another area. In counties with relatively fewer opportunities to spend money, expenditures based on nights spent will likely be overestimated. On the other hand, expenditures in counties with relatively more spending opportunities will likely be underestimated. Also, using the percent of nights spent in a county ignores day trippers’ expenditures and expenditures made on route to the places where people spend their nights. These expenditures can influence the amount of money spent in an area and the share of money each area receives.

There is also a limitation associated with the data on where visitors spent nights. During the summer and shoulder seasons, visitors only reported the location where the stayed during the first six nights of their trip (12 nights were reported in the winter study). Although the average number of nights visitors stay in the state is less than four, the fact that not all nights were reported, nor were these nights randomly selected, means that the data could be somewhat biased.

Finally, there is one indication that the figures in Table 1 are overestimated. Based on the accommodations tax, Madison and Lincoln Counties each have about $1.8 million in lodging sales (total tax collections were about $73,000 in each of the counties in 1997, which is approximately 4% of $1.8 million). If one applies the share of visitors’ expenditures that were spent on lodging in each county (Table 3), to the estimates of total expenditures based on nights (Table 1), the results are estimates of over $2 million (8% of $26 million) in lodging expenditures by nonresident visitors for Lincoln County and $3 million (17% of $18.6 million) for Madison County. These amounts are both higher than the estimated total lodging sales in the counties.

Total expenditures in counties estimated by percent spending in counties (Table 2)

Having actual spending data for an area is preferred over trying to estimate spending based on some other indicator. Using actual spending data eliminates the difficulties in accounting for different spending patterns, prices, type of visitor (e.g., day versus overnight visitors) and accommodations (e.g., staying with friends and relatives) associated with other methods. Therefore, the method used to arrive at the estimates in Table 2 is considered the preferred method, provided there is enough information contained in the surveys.

The limitation associated with estimating county-level expenditures is primarily a lack of data, particularly for the winter and shoulder seasons and for places that receive relatively less visitation. Enough surveys are collected to provide reliable estimates for the state as a whole. As one tries to break the data down further by county or city, one quickly runs into difficulty because of a lack of observations. Also, items that are bought less often (e.g., transportation and auto repairs) are more likely to be missed as the data are broken down further, leaving us with little or no information regarding these types of purchases.

Finally, there is one indication that the figures in Table 2 are reasonable. If one applies the share of visitors’ expenditures that were spent on lodging in each county (Table 3), to the estimates of total expenditures based on spending in the counties (Table 2), the results are over $1 million (8% of $14.7 million) in lodging expenditures by nonresident visitors for Lincoln County and $1.5 million (17% of $9 million) for Madison County. These amounts are both reasonable in comparison to the estimated amount of total lodging sales in the two counties ($1.8 million in each county, based on the accommodations tax). The difference between the estimated amount spent by nonresident visitors and the total lodging sales can be attributed to resident, commercial, and other use not included in ITRR surveys (plus a margin of error in the nonresident estimate).

Note in most cases that the percent of money spent in each county is less than the percent of nights spent in the counties. This difference is probably a function of relatively fewer opportunities to spend money in these places and/or relatively less costly goods and services (relative to the overall state picture).

Comparisons with the accommodations tax

The fact that the percent of nights spent in each county (1.7% in Lincoln and 1.2% in Madison) is greater than the percent of accommodations tax paid in each county (0.8% each) suggests a couple of possible differences between these two counties and the state as a whole. First, the number of nights spent in federal campgrounds (or backcountry camping) or with friends and relatives might be higher than statewide average. Second, the room rates in these counties could be lower than the statewide average.

General observation

The fact that each of the estimates are relatively similar in terms of total spending in the state (0.6% to 1.7% for both counties) is encouraging. But, when applying these percentages to $1.5 billion, these seemingly minor differences can be significant for a small county.

Conclusions

Estimating visitor expenditures for counties is difficult, but possible in some cases. It is preferable to use observations of actual spending in counties, rather than basing estimates on an indicator like the accommodations tax or number of nights spent in a county. ITRR survey data can be used in some cases to estimate county-level expenditures. The main problem is a lack of data for areas that receive relatively small amounts of visitation. The less visited a county is, the less reliable data for that county will be, and vice-versa.

In addition, due to the enormous effort required to carry out a yearlong study and the limited resources available to do so, ITRR’s most recent nonresident studies have been seasonal. Therefore, estimating expenditures for an entire year using the most recent data available requires piecing together data from three separate databases, each of which is slightly different. This presents a challenge simply in the amount of time and effort required to retrieve data for a county.

Sources