The Impact of Sporting and Cultural Events in a Heterogeneous Hotel Market: Evidence from Austin, TX

Abstract

This paper provides an empirical analysis of how various local political, cultural, and sporting events affect daily room rentals in the Austin, Texas hotel market. In an innovation of previous studies that use daily hotel data, we differentiate how events affect different tiers of hotel rooms ranging from economy hotels to luxury hotels. The results show disparities between events with regard to the quality of hotel demanded. The analysis is extended to daily hotel revenues generated in the six tiers of hotel rooms and how the various events contribute to local and state hotel-occupancy tax revenues.

Introduction

Local politicians and tourism promoters often point to tourist inflows associated with sporting and cultural events to justify public subsidies for the events. However, more than three decades of research on the economic impact of sporting events find scant evidence that hosting such events generates benefits commensurate with their public subsidies.¹ Of course, overpromised benefits are just one aspect of what Müller (2015) refers to as “mega-event syndrome.”²

Although the research examined by Coates and Humphreys (2008) was nearly unanimous in finding little benefit from sports franchises, stadiums, and mega-events, research on this topic has continued. Among the new approaches are the use of granular data measured near where the events are held and in ever-shorter time periods. Examples include analyzing monthly sales tax revenues (Baade and Matheson 2001; Baade et al. 2008; and Coates and Depken 2011), daily airline passenger traffic (Baumann et al. 2009; and Baumann and Matheson 2016), daily crime rates (Billings and Depken 2011), and daily restaurant revenue (Depken and Fore 2020).

Hotel occupancy data have also been analyzed to measure events’ economic benefits. Since the local economic impact of hosting a large event consists primarily of increased spending by visitors from outside the host city or region,³ hotel occupancy data are well suited for analyzing the ability of various events to attract visitors. Lavoie and Rodriguez (2005) analyze monthly hotel occupancy data from eight large Canadian cities and find little evidence that NBA and NHL franchises attract large numbers of overnight visitors.

More recently, daily hotel occupancy data have become available and have led to several papers examining sports or other events.⁴ Collins and Stephenson (2016) analyze the effect of the National Association of Intercollegiate Athletics (NAIA) college football national championship on Rome, Georgia. They find a small positive economic impact of the event, including some evidence of guests arriving a few days before the game day, but little evidence of guests staying after the game day.

Depken and Stephenson (2018) use daily hotel occupancy data to examine the effects of NFL games, NBA games, NASCAR races, several college sports competitions, and major conventions in Charlotte, North Carolina. They show that effects vary considerably across events, with relatively large effects associated with NASCAR races and a regional college basketball tournament, modest effects arising from NFL games, no effect resulting from NBA games, and substantial effects associated with conventions. The lead and lag variables included in their analysis provide little evidence of intertemporal spillover benefits for most events studied.

Similarly, Heller et al. (2018) use daily hotel occupancy data to analyze the four national political party conventions held in 2008 and 2012. Host cities often claim large economic benefits from hosting political conventions. Heller et al. (2018) do find that conventions produce substantial bumps in hotel occupancy, but they point out that the effects are considerably smaller than host cities often claim. Using leads and lags before and after political conventions, Heller et al. (2018) also identify a “hangover effect” of reduced hotel occupancy immediately following political conventions. Such effects might be caused by the need to clean an arena or convention center after an event and should be incorporated into the overall assessment of economic impact associated with an event.

Chikish et al. (2019) investigate the effect of basketball and hockey events held in the Staples Center in Los Angeles on local daily hotel demand. They find that the impact of events is highly concentrated around the Staples Center. Since hotels in the immediate vicinity of the Staples Center are exempt from hotel occupancy taxes, Los Angeles is forgoing substantial hotel-occupancy tax revenue that would be generated by professional basketball and hockey games.

Using a similar approach to Depken and Stephenson (2018), Heller and Stephenson (2021) examine the Super Bowl’s effects on daily hotel occupancy in four host cities. Their key findings include that the effect varies across cities depending on “normal” hotel occupancy during the time of the year that the Super Bowl is played and that the Super Bowl tourist inflow may go to other parts of a region than the host city.

Daily hotel occupancy data offer several advantages over monthly hotel occupancy data. First, daily data can better control for both time of year and day of week effects, thereby allowing the estimation of net rather than gross effects associated with the sports events. Obtaining the net effect requires capturing normal time of year and day of week occupancy rates to control for the crowding out of normal hotel occupancy (Porter 1999). Second, as indicated by descriptions of previous research using daily hotel occupancy data, daily data can incorporate variables for the days leading up to or following an event to better capture the duration of visitors’ stays. Event promoters sometimes suggest that hosting a large event such as the Super Bowl entices visitors to arrive several days before the event or stay several days after the event; the use of leads and lags can capture such effects. Third, daily data are cleaner than using month dummies when studying events such as World Cup tournaments, which normally span multiple months. Fourth, daily data might be able to detect small effects that would be lost in monthly data. For example, consider a one-day event that generates 2000 additional room nights. If the host city averages 20,000 rooms let per night, the 2000-room increase might be detectable using daily data because it would reflect a 10% increase in rooms let. However, the same 2,000-room night increase would be difficult to detect using monthly data since the city would average approximately 600,000 room nights per month (= 30 * 20,000).

In this paper, we further refine recent research by examining the effect of various sporting events on the different tiers of hotels (e.g., economy vs. luxury).⁵ Understanding whether events have heterogeneous effects across hotel types might be important for tourist bureaus attempting to attract the best events for their city. Knowing that an event tends to fill more hotels that are expensive could be valuable information when considering the costs and benefits of hosting an event. One reason is that hotel taxes are ad valorem, that is, assessed as a percentage of the nightly room rate. Thus, knowing which events increase occupancy at more expensive hotels provides information about which events generate the largest increase in hotel tax receipts. Another reason is that willingness to spend on hotels is positively correlated with willingness to spend on other things like bar tabs, restaurant meals, souvenirs, etc. (Marcussen 2011). Hence, events that yield more room rentals in high-end hotels likely also generate larger increases in visitor spending on other goods and services.

Data and Estimation Strategy

The data employed in this empirical study describe the daily hotel market for Austin, Texas, from January 1, 2010, through February 28, 2017. For each day, the data contain the total number of rooms rented and the total hotel revenue from room rentals. The hotel data cover the metro region of Austin, Texas, and were obtained from STR, a clearinghouse for hotel market data in the USA and overseas. Since our analysis focuses on hotel occupancy data, it is worth noting upfront a couple of limitations to our study. First, our study does not capture any economic gains from day visitors who do not stay overnight. Second, our study does not capture any overnight visitors who opt for lodging other than hotels such as short-term rentals (e.g., AirBnB) or staying with local friends or family.

The estimation strategy relates the daily hotel rooms or daily revenue (${DEP}_{\ell{t}}$) in each of $\ell=6$ tiers of hotel quality to indicator variables that take a value of one on days during which specific events are held and zero otherwise (EVENT_it), and additional control variables. The estimating equation is as follows:

$${DEP}_{\ell{t}}= {\beta }_{0}+\sum_{i=1}^{M}{\beta }_{i}{EVENT}_{it}+{\sum_{j=1}^{{J}_{i}}{{\theta }_{j}EVENT}_{it-j}+\sum_{k=1}^{{K}_{i}}{{\varphi }_{k}EVENT}_{it+k}+\delta }_{1}{GASPRICE}_{t}+{\delta }_{2}U{NEMPLOYMENT}_{t}+ {\gamma }_{1}DAY+{ \gamma }_{2}MONTH+{\gamma }_{3}YEAR+{\varepsilon }_{t},$$

(1)

where ${DEP}_{\ell{t}}$ is, alternatively, the number of rooms rented or total revenue in hotel tier $\ell$ in time t, the $\beta$’s, $\theta$’s, $\varphi$’s, $\delta$’s, and $\gamma$'s are parameters to be estimated, and ε is a zero mean error term.⁶ The $\ell$ tiers of hotels are Luxury, Upper High Tier, High Tier, Upper Middle Tier, Middle Tier, and Economy.⁷ To control for general macroeconomic conditions that might influence the hotel market, we include the real gasoline price for the week and the month's unemployment rate. Following Depken and Stephenson (2018), we model how events impact daily hotel room rentals during a multi-day window centered on the event day(s), that is, J_i=2 and K_i =2. Augmented Dickey–Fuller tests indicated that the data are stationary so the model is estimated via OLS with Newey–West corrected standard errors to control for serial correlation.⁸

The M=6 events included in our analysis and the number of events are reported in Table 1. The events include college football games hosted at the University of Texas at Austin (UT), commencement ceremonies held at UT, the South-by-Southwest Music Festival (SXSW), the Summer X-Games, the Six Hours of COTA (a session of the World Endurance Competition), Motorcycle Grand Prix (MotoGP) races, and Formula One automobile races (the last three events are held at Circuit of the Americas, a race track located near Austin). Among the events modeled, there are 7 SXSW music festivals, 74 college football games, 12 college commencement ceremonies, 3 Summer X-Games, 4 Six Hours of COTA, 4 MotoGP races, and 5 Formula One races.

Table 1 Events analyzed

Table 2 reports descriptive statistics for the primary variables in the model along with the two exogenous control variables. From the descriptive statistics concerning the tiers of hotels, it is apparent that Upper High Tier, High Tier, and Economy hotel rooms are the most commonly rented on a given day and that these three tiers also have some of the highest maximums and minimums over the sample period. By contrast, rooms at Luxury hotels are the least commonly rented. The substantial variation in the number of rooms rented over the sample period invites further investigation as to how various events influence the daily hotel market and whether the events disparately influence the room rentals for various tiers of hotels.

Table 2 Descriptive statistics of the sample

Estimation Results

Hotel Rooms Rented

To investigate the impact of the various events on daily hotel rooms rented, we estimated the model separately for each of the six tiers. Tables 3 through 6 report the estimation results for the six models with each row reflecting a different tier of hotel and the various columns corresponding to a window centered on an event. Since UT football games, UT commencements, Six Hours of COTA, and MotoGP are all single-day events, the estimation window for them is comprised of five days, the event day along with two days before and two days after the event. For SXSW, the event window is 14 days, the 10 festival days plus two days before and two days after. For the X-Games, the event window is 8 days, the four-day competition plus two days before and two days after. For Formula One races, the event window is 7 days, the 3 race days plus two days before and two days after.

Table 3 Impact of events on daily room rentals at various hotel tiers

From Table 3, we see that SXSW has a large and consistent impact on the daily hotel market in the Austin metro area. In every tier of hotel and in every day of the fourteen-day analysis window, there is a positive and statistically significant increase in the number of rooms let. The largest increase occurs in Economy hotels—more than 1,000 additional rooms let on average over the ten-day festival as well as the two nights leading up SXSW and the night after its conclusion. The estimated effect of SXSW tends to decrease, though not monotonically, as the hotel tier increases, culminating in a modest increase of about 75 additional Luxury tier rooms let during each night of the festival. This pattern may indicate that many music festival attendees are young and less affluent. Regardless of tier, the results for SXSW are consistent with common claims that big events can induce pre-event and post-event tourism effects.

Table 3 also reports the estimation results for the Summer X-Games. While the X-Games, like SXSW, are a multi-day event, the impact of the X-Games on the daily hotel market in Austin is considerably smaller. Indeed, there is only an increase in the number of rooms rented after the event is over. The insignificant effects on hotel room rentals in every hotel tier before and during the event suggest that all those who are renting hotel rooms to visit the event are substituting for those who would have been renting rooms if the X-Games were not occurring.⁹ The increase in rooms let after the event suggests local business activity may have been postponed until after the event.

Table 4 reports the results for events associated with UT. In the case of UT football games, all tiers have a statistically significant increase on game days. The largest increase is observed for Upper Middle Tier hotels, though the increases for High tier and Upper High tier are nearly as large. The increases for Economy and Middle Tier hotels are somewhat smaller than the Upper Middle, High, and Upper High Tiers, which is consistent with perceptions that many out-of-town game attendees are older alumni who are more established financially. Across all tiers, the total effect on event day is fewer than 3000 rooms let, which indicates that the vast majority of the 90,000 fans who attend a typical game do not stay overnight in hotels.

Table 4 Impact of events on daily room rentals in various hotel tiers

As for the days before and after UT football games, there is no statistically significant effect one day before a game or one day after a game. However, there is a reduction in the number of rooms rented in each hotel tier two days before a game and two days after a game. Since most UT games are on Saturday, the decrease in room rentals two days before and after games indicates that rentals are lower on Thursdays and Mondays. However, the models already control for day of the week effects, so these decreases would apply only to Thursdays and Mondays adjacent to football home games. A possible explanation for the decreased room rentals two days before and after UT football games is that UT games might discourage business activity, perhaps out of concerns about increased congestion or increases in complementary travel costs such as airfare.

From Table 4, the day of UT commencement ceremonies is associated with a statistically significant increase in the rooms rented in all tiers except Luxury hotels, with Economy hotels seeing a somewhat larger increase than any other tier. Most tiers also see statistically significant increases both one day and two days before the commencement ceremonies, but no tiers see statistically significant increases one day after or two days after the event. This pattern of increased rentals leading up to the commencement but no increase afterwards is not surprising, since many universities have social events in the days preceding their commencement ceremonies.

From Table 5, the Six Hours of COTA event is associated with statistically significant increases in rooms rented in all but Upper High tier hotels on the day of the event. High Tier, Upper Middle Tier, and Economy hotels all have increases of about 300 rooms let, while Luxury and Middle Tier hotels have increases roughly half as large. These results suggest the endurance event attracts an economically diverse audience. Most tiers see statistically significant increases in the days leading up to the event but show little evidence of fans remaining in Austin beyond the event.

Table 5 Impact of events on daily room rentals in various hotel tiers

As for MotoGP events, Table 5 reports that there are statistically significant increases in room rentals in all hotel tiers except Luxury on the night of the event. The pattern is nearly the opposite of that observed for SXSW: The largest increase is in Upper High tier room rentals and the effect decreases, though not monotonically, as one moves down toward the Economy tier. Like Six Hours of COTA, both one and two days before the event have large increases in rooms let, whereas there is scant evidence of increased room rentals one and two days after the event.

Table 6 Impact of events on daily room rentals in various hotel tiers

Table 6 reports the results concerning the Formula One race held at the Circuit of the Americas. This event is one of the premier events hosted in the Austin region. The Circuit of the Americas was built with a combination of private and public money with the intention of recruiting and retaining the only Formula One race held in the USA. As such, it is expected that the Formula One race would have a large impact on the daily hotel market in Austin. Furthermore, the fan base of Formula One is often described as upper middle to upper class people attracted to high-quality high-priced consumer products.¹⁰ Some of this preference is reflected in the corporate sponsors of Formula One compared to NASCAR and other racing circuits in the USA.¹¹

As with SXSW, the Formula One event days are associated with statistically significant increases in rooms rented in all tiers of hotels. Consistent with the perception that Formula One attracts well-heeled fans, the increase in luxury room rentals is larger than any other event included in this study, more than triple Six Hours of COTA, which has the next largest effect. Formula One’s effect on Luxury hotel room rentals is nearly as large as its effect on Economy and Middle Tier room rentals, even though there are many fewer Luxury rooms available. As with MotoGP and Six Hours of COTA, Formula One is also associated with large increases in room rentals in the two days prior to the event but has little effect in the days after the event.

Hotel Room Revenue

In this section, we repeat the estimation replacing the number of rooms rented in each hotel tier as the dependent variable with total room revenue in each hotel tier. As with the room rental results discussed in the previous section, the model is estimated once for each hotel tier and the reported results are grouped by event. Tables 7 through 10 report the estimation results for the six models, with each row reflecting a different hotel tier and the various columns corresponding to the event windows. As with the room rental estimation, all models contain day of the week and time of year fixed effects, the price of gasoline, and the unemployment rate.

From Table 7, the impact of SXSW on daily hotel revenue is positive and statistically significant for all tiers and each day of the fourteen-day window. The increased revenue from High Tier, Upper High Tier, and Luxury rooms is much greater than the increased revenues generated in the lower three tiers. Thus, the music festival generates considerable tax revenues for the city of Austin and the state of Texas and represents the greatest revenue increase among the events studied here.

In contrast, Table 7 shows there is no statistically significant increase in revenues in any hotel tier two days before the Summer X-Games. On the day before the X-Games event, revenues increase slightly for Luxury, Middle Tier, and Economy hotels. On the four X-Games event days, revenues are slightly higher for Upper Middle Tier, Middle Tier, and Economy hotels. One day and two days after the X-Games, revenues are higher in all hotel tiers. Overall, there appears to be a net positive impact on hotel revenues and, therefore, on hotel-related taxes to the city of Austin and the state of Texas from hosting the Summer X-Games. However, the impact is considerably smaller than that of SXSW both because the X-Games last fewer nights and because they generate smaller gains in hotel revenue per night.

From Table 8, two days before a UT home football game there are statistically significant reductions in total revenue in all tiers other than Luxury and Economy hotels. One day before football games, there are statistically significant increases in total revenue in all tiers other than High Tier hotels. On the day of the game, there are statistically significant and substantial increases in revenues in all hotel tiers, with the largest gains being in Upper High Tier hotels, High Tier hotels, and Upper Middle Tier hotels. The day after the game, there are no statistically significant changes in revenues in any hotel tier, and there are statistically significant decreases in revenue for all hotel tiers two days after a UT football game.

Table 7 Impact of events on daily revenue at various hotel tiers

Table 8 Impact of events on daily revenue in various hotel tiers

As shown in Table 8, the two days before, the day before, and the day of UT commencement ceremonies correspond with statistically significant and substantial increases in revenues in all hotel tiers, but the increases are noticeably greater in the higher tiers. There is no statistically significant change in hotel revenues in the days after UT commencement ceremonies compared to other similar days of the month that do not have a ceremony associated with them.

Table 9 reports the estimation results of how daily hotel revenue is impacted by motorsport events held at the Circuit of the Americas. The Six Hours at COTA endurance race is associated with an increase in daily revenue for all tiers of hotels two days before, one day before, and day of the event. There is no statistically significant impact on hotel revenues the day after the event, and only one tier, High Tier, shows an increase in revenue two days after the event. As is the case in other events, the increase in revenue is more noticeable in the higher tiers of hotel rooms, suggesting that the individuals who rent these types of rooms pay more in hotel-occupancy tax.

Table 9 Impact of events on daily revenue in various hotel tiers

Table 9 also reports the estimation results for the MotoGP. Again, there is an increase in hotel revenue for all tiers two days before and one day before the event. There is an increase in revenue for all tiers other than Luxury rooms on the day of the event. There is a slight decline in revenues the day after and two days after the event in Economy hotels; there is no change in daily revenue in other tiers the day after or two days after the event.

From Table 10, the impact of the Formula One race on daily hotel revenue is positive and statistically significant for all tiers two days before, one day before, and day of the event. The increased revenues in the High Tier, Upper High Tier, and Luxury tiers are all considerably larger than those obtained in the Middle Tier and Economy Tiers. The revenue impacts in Luxury rooms, Upper High Tier, and High Tier rooms are the largest estimated for any of the events included in this study and the impact in Upper High Tier is the only estimate to exceed $1m per day in additional revenue. This suggests that the customer base that is drawn to Formula One races is skewed towards Luxury, Upper High Tier, and High Tier rooms. While there are increases in revenue in the Middle and Lower Tier, the increases in the other tiers are substantially larger. There are no statistically significant effects on hotel revenues following the race except for a reduction in Economy hotel revenue two days after the event.

Table 10 Impact of events on daily revenue in various hotel tiers

Discussion and Policy Implications

The city of Austin imposes an eleven percent tax on hotel room rentals. Nine percentage points are an occupancy tax, remitted to the general fund for the city of Austin, and two percentage points are dedicated to the city’s venue construction and event hosting fund. The state of Texas imposes an additional six percent tax on hotel room rentals (AustinTexas.gov 2019). Therefore, there are direct policy implications as to how different events influence the market for the various hotel tiers because of the tax implications. Consider an event that has a large impact on Economy hotels. Because of the generally lower room rates at these properties, the tax revenues are relatively low and might fall below the costs of hosting the event. On the other hand, if an event has a large impact on Upper High Tier and Luxury rooms, which have higher room rates, then the tax payments to the city and state might be large and could offset the costs of hosting the events.

Table 11 reports aggregate hotel room revenue during each event’s window using only those parameters from Tables 7–10 that have a p ≤ 0.10. Table 11 also reports the estimated hotel occupancy tax dedicated to Austin’s general fund, the portion dedicated to the city of Austin’s venue construction and event hosting fund, and the estimated sales tax remitted to the state of Texas. The last row of Table 11 reports Grand Totals from a representative year of hosting events in the city of Austin to include 1 ten-day SXSW music festival, 6 UT football games, 2 UT commencement ceremonies, 1 Six Hours of COTA event, 1 MotoGP event, 1 three-day Formula One event, and 1 four-day Summer X-Games event. (Recall, however, from Table 1 that Austin did not host all events each year of the study period.)

Table 11 Estimated total net revenue and tax revenue impacts of multi-day event windows

From Table 11, the net impact on hotel revenues during SXSW is approximately $30.1m, primarily because the event lasts ten days and has a significant impact on all tiers of hotels. This increase in hotel revenues corresponds with a $2.7m increase in the city of Austin’s general fund, approximately $0.6m to the Austin events fund, and approximately $1.8m to the state of Texas.

The four-day X-Games correspond with an increase in overall hotel revenues of approximately $1.45m, primarily because the X-Games do not have a statistically significant impact on several tiers of hotels, but do contribute $131k, $29k, and $87.1k, to the general Austin fund, the Austin events fund, and the state of Texas, respectively.

Six UT football games increase total hotel revenues by $1.90m, primarily because of the offsetting relatively large increases in hotel revenues the day before and day of the game and relatively large decreases in hotel revenue two days before and two days after the events. As a result, UT football games contribute $170.6k, $37.9k, and $113.7k to the city of Austin, the city of Austin’s event fund, and the state of Texas, respectively. While there are large and obvious increases in hotel revenues that occur the day before and the day of UT football games, this analysis shows there are regular decreases in revenues in the days before and after the football games which are not obvious to casual observation.

The net impact of UT commencement exercises on hotel revenue is substantial at $4.33m. The difference between football games and commencements lies in the revenue increases in the days leading up to commencements and the lack of decreased revenues afterward. This suggests that multi-day visits before commencements are more common for commencements relative to football games. This might reflect the inherent differences in the characteristics and audiences of the two different events at UT. UT commencements add $389.8k, $86.6k, and $259.9k, to the city of Austin, the city of Austin’s event fund, and the state of Texas, respectively.

The events at the Circuit of the Americas racetrack have positive total net impacts on hotel revenues. The Six Hours of COTA has a net impact of $2.13m, the Motorcycle Grand Prix has a net impact of $2.82m, and the Formula One race has a net impact of $20.29m. As expected, Formula One has the greatest effect on total hotel revenue of the races analyzed because of the large increases in revenue at Luxury, Upper High, and High tier hotels in the days leading up to and including the race. Table 11 shows that while all three races contribute positively to the city of Austin, the Austin event fund, and the state of Texas, the Formula One race exceeds the other two races nearly eight-fold.

A representative year of hosting events would lead to an estimated net increase in hotel revenues of approximately $63.02m, an increase in taxes dedicated to the Austin city general fund of $5.67m, an increase in taxes dedicated to the Austin event fund of $1.26m, and an increase in taxes remitted to the state of Texas of $3.78m. These estimated tax implications of hosting events entail only changes in hotel room revenue and do not include taxes gathered on other tourism-related spending, such as food and beverages, or other spending subject to state and local sales taxes.

Knowing more about how these various events generate additional hotel revenues, especially the events hosted at the Circuit of the Americas, also has policy implications in other dimensions. The state of Texas has a Major Events Reimbursement Program (MERP), formally known as the Major Events Trust Fund (METF). Host cities and local event organizing committees in the state of Texas can apply for reimbursement of costs associated with attracting and hosting events. Applications must be submitted before the event and require an economic impact study that predicts the amount of additional economic activity through direct spending, indirect spending, and induced spending associated with the event. Once the event occurs and the actual attendance is verified, a reimbursement to the host city or local organizing committee is made. The funds are financed by a portion of the additional revenues generated by the event (and up to ten months afterward) through the general sales tax, any hotel occupancy tax, any vehicle rental tax, any mixed beverage tax, and the wholesale alcohol tax. The state matches local contributions at the rate of 6.5:1. If the predictions of the event’s economic impact or attendance is substantially different from what occurs, the event can be disqualified from future participation in the program.

When plans to build the COTA were initially discussed in 2010, one of the primary goals was to recruit a Formula One race. However, it became apparent that Formula One was going to charge an annual sanctioning fee of $25m to COTA for the rights to host a Formula One race. It was also apparent that it was unlikely that the COTA and its corporate sponsors would be able to pay the sanctioning fee without public assistance.

The MERP, then called the METF, was used to reimburse a local organizing committee for the sanctioning fee, generating substantial controversy centered on this large transfer to Formula One, a private entity that is not based in the USA. The economic impact study that accompanies the annual application of the COTA for reimbursement of the sanctioning fee asserts that the Formula One race is associated with a total economic impact of over $400m and that the state of Texas receives more than $50m per year in additional tax revenue generated by the race.

However, the results herein show that the assumption that all hotel spending is the same and subject to the same rate of leakage from the local economy is likely incorrect. The distinction in this study is the use of granular data that distinguishes between tiers of hotels and when hotel rooms are rented relative to event days. If Luxury hotels are more likely to be owned by national or international chains, more of the revenue generated in these types of hotels will leak from the local economy and therefore will not be included in additional rounds of induced local spending. If, by contrast, Economy hotels are more likely to be owned by local residents, then the additional revenue generated at these hotels is more likely to remain in the local economy and contribute to induced spending. Hence, our results show that the impacts of events on hotel revenues and attendant taxes are nuanced, which should be incorporated into future economic impact studies.

Conclusions

Depken and Stephenson (2018) differentiate hotel markets by inclusion in local hotel-tax jurisdictions and distance from venues in Charlotte, North Carolina. Chikish et al. (2019) differentiate hotel markets by their distance to Staples Center in Los Angeles and inclusion in tax-exemption zones. Here, we segment the hotel market by the tier of hotel. This provides a different context for how various events influence heterogeneous hotel markets.

We analyze the impact of seven regularly hosted events in Austin, Texas, on daily room rentals and daily revenue, focusing on a multi-day window centered on the event day. Surprisingly, UT football games have a net negative impact on rooms rented (and total revenue) once considering the days before and days after the day of the game. All other events have a net positive impact on rooms rented and total revenue, though in some hotel tiers there is a net decline in rooms rented during the multi-day window. Thus, we conclude that the events included in this study have different impacts in a heterogeneous hotel market. Some events seem to draw tourists who prefer to stay in Economy hotels, while other events draw tourists who prefer to stay in Luxury hotels.

We extend the analysis to a discussion of the impact of events on local and state tax revenues. Austin’s hotel-occupancy tax revenue and the state of Texas sales tax revenue increase for six of the seven events included in this study. What is less clear is the relationship between increased spending on accommodations and net economic activity in Austin associated with the events. The literature focusing on the relationship between accommodation spending and other spending during a tourism stay is sparse, though Marcussen (2011) finds a positive relationship.

However, such considerations are important for determining the expected net economic impact on the local economy of hosting an event and the total amount of additional taxes available to reimburse the MERP. To date, many aspects of hosting sporting events, including sales taxes, wages, jobs, property values, and crime patterns, have been investigated.¹² Recently, daily hotel markets have become an added dimension in which to analyze the impact of hosted events. The next leap in data availability would be granular data at the consumer level, including hotel spending and other spending in the local hospitality industry. Such data would facilitate investigation of substitution and complementarity between accommodation spending and other spending in the hospitality industry.

Lastly, this study also helps understand the economic harm associated with the Covid-19 pandemic. One of the first high-profile event cancelations at the beginning of the pandemic was Austin’s announcement on March 6, 2020 that SXSW, slated to start later that month, would not be held. This paper’s results shed some light on the economic harm associated with not holding the music festival and subsequent cancellations during the 2020 and 2021 pandemic.