Economic multipliers and mega-event analysis.

Matheson, Victor A.

Introduction

Economists often criticize economic impact studies that purport to show that mega-events such as the Olympics, World Cup, or other sports championships such as the Super Bowl bring large benefits to the communities "lucky" enough to host them. These scholars frequently cite the use of inappropriate multipliers as one of the primary reasons why these impact studies overstate the economic gains to the hosts of these events (see Siegfried and Zimbalist (2000), Crompton (1995), or Baade and Matheson (2001) among others). Porter and Fletcher (2008) echo these concerns, going so far as to note that most academic journals will not publish economic impact studies generated using the most commonly used software packages.

The concept of multipliers is well established in the field of economics, however, and indeed the 1973 Nobel Prize in Economics was awarded to Wassily Leontief for his work in developing the macroeconomic input-output models used to derive multipliers. Therefore, it is not appropriate to simply reject, out of hand, all use of multipliers in mega-event impact analysis without a solid economic reason for doing so. The purpose of this paper is to demonstrate one justification for rejecting the use of standard economic multipliers in the analysis of the economic impact of mega-events.

Before proceeding further, it is important to more precisely define the terminology economists use when discussing multipliers. First, practitioners of economic impact analysis are often quite vague about the distinctions between economic impact, economic benefits, and increased spending and typically use these terms interchangeably. Of course, increased spending in an area does not necessarily lead to increased incomes, and an economy may not "benefit" in an appreciable way just because spending increases. While economists would most likely define the benefits of an event to a city as being related to the income generated for its citizens, economic impact reports invariably equate economic impact with spending, and, therefore, it is this that will be examined in this paper.

Economists also use two differing conventions in reporting multipliers. One method calculates the multiplier as equal to indirect spending divided by direct spending, so that a multiplier of 1 results in total spending being double that of the direct spending. Others, such as Humphreys (1994), report that the multiplier equals indirect spending plus direct spending all divided by direct spending, so that instead a multiplier of 2 implies a doubling of direct spending. The second convention seems more natural and more widespread, so it will be used in the remainder of this paper.

Economic impact analysis is generally done by estimating attendance at an event, surveying a sample of visitors as to their spending associated with the game or convention, and then applying a multiplier to account for money circulating through the economy after the initial round of spending. For example, an economic impact analysis for the American football championship game, Super Bowl XXVIII, in Atlanta in 1994 estimated 306,680 visitor days with a typical visitor spending $252 per day for a direct impact of $77.3 million. An economic multiplier of 2.148 is then applied for an indirect impact of $88.7 million and a total economic benefit of $166 million (Humphreys, 1994).

The economic multipliers used in these analyses are calculated using complex input-output tables for specific industries. One commonly used model in the United States is the Bureau of Economic Analysis' Regional Industrial Multiplier System (RIMS II) that provides final-demand output multipliers for 473 detailed industries, including hotel accommodations, eating and drinking establishments, and arts, entertainment, and recreation. One common criticism of the use of multipliers in the analysis of professional sports is simply that the multipliers used are too high since many athletes live outside the local area in which they play. Therefore, wages paid to these athletes are less likely to recirculate throughout the local economy than wages paid to workers in other fields leading to a lower multiplier effect for professional sports than in other industries (Siegfried & Zimbalist, 2002). While this criticism is certainly valid, the problem can be addressed simply by creating input-output tables at a sufficient level of detail to specifically address the peculiarities of the professional sports industry.

When examining mega-events, however, the problem of inflated multipliers becomes even more problematic. The multipliers in RIMS II (or other multiplier models) are based upon inter-industry relationships within regions based upon an economic area's normal production patterns. During mega-events, however, the economy within a region may be anything but normal, and therefore, these same inter-industry relationships may not hold. As noted by Porter and Fletcher (2008), since there is no reason to believe that the usual economic multipliers are the same during mega-events, any economic analyses based upon these multipliers may be highly inaccurate. Indeed, there is substantial reason to believe that during mega-events, these multipliers are highly overstated, and, therefore, their use overestimates the true impact of these events on the local economy. This concept is easily explored through a simple numerical example.

Numerical Example

Suppose the hotel industry, an industry that accounts for up to half of all visitor spending during mega-events, is characterized by a situation where hotel service is provided by combining capital, which can be supplied either locally or by national or international capital markets, and labor, which is supplied exclusively by local workers. Income earned by capital owners (or stockholders) who do not live in the city in which the hotel is located is unlikely to be respent in the local economy in comparison to wages earned by local labor. Revenues that flow out of an economy after an initial round of spending are typically referred to as "leakages."

In particular, suppose that a hotel typically supplies 75 rooms at a price of $150 per night requiring the use of 75 workers earning $100 per day. Any revenues in excess of labor costs accrue to capital owners as profit. The multiplier effect from hotel expenditures depends on how labor and capital spend their respective earnings in the local economy. If both the laborers and hotel owners are local, assume that 50% of their earnings are re-spent on local goods and that a multiplier equal to 2 is applied to any subsequent rounds of spending. The direct and indirect impact of hotel spending is shown in Table 1, Scenario 1.1 Alternatively, if the hotel is part of a nationally owned chain, the workers are still local, but the capital owners are national, and, therefore, income earned by capital will not recirculate in the economy as shown in Table 1, Scenario 1.2.

Scenario 2

Suppose the mega-event increases the number of rooms sold to 100 while leaving the room prices unchanged. The hotel hires 100 workers (instead of 75) in order to accommodate the higher demand. This additional labor can be drawn from the existing labor pool by offering more hours to existing workers or by tapping into the ranks of the unemployed. Either way, earnings to labor increases with the event. The corresponding direct and indirect earnings from the event if the hotel is locally owned is shown in Table 1, Scenario 2.1, while the figures for a nationally owned hotel are shown in Scenario 2.2. As seen in the table, as long as the higher demand from the mega-event results in labor and capital equally sharing in the increased revenue, then the corresponding multipliers remain unchanged.

A second important fact can be shown from comparing Scenarios 1.1 and 1.2 to Scenarios 2.1 and 2.2. Even if all hotel revenues during a specific period can be attributed to a particular event, while the gross hotel revenues associated with the event are high, the marginal revenues are much lower because the event visitors crowd out the regular hotel business. For example, in Scenario 2.1, the gross direct hotel revenues are 15,000; however, the net direct impact of the event is only the difference between hotel revenues during the event and revenues typically, or 3,750.

Scenario 3

Suppose the mega-event increases the number of rooms sold to 100 while leaving the room prices unchanged. In this scenario, however, the hotel does not hire additional workers in the face of the higher demand. The existing workers are simply expected to work harder or more efficiently to order to meet the customers' needs in the crowded hotel. The corresponding direct and indirect earnings from the event if the hotel is locally owned is shown in Table 1, Scenario 3.1, while the figures for a nationally owned hotel are shown in Scenario 3.2. As seen in the table, when capital and labor are both locally supplied, the multiplier is unaffected by the distribution of the proceeds from the event among the factors of production since both will re-spend the same fraction of their earnings locally. When the hotel is nationally owned, however, the increase in the portion of the hotel's revenue that accrues to capital serves to reduce the multiplier. Indeed, the mega-event results in no marginal increase in indirect spending whatsoever. A typical impact analysis would apply the usual multiplier of 1.67 to the 15,000 in direct hotel spending to arrive at an estimate of a 25,000 (= 1.67 x 15,000 = 25,000) gain from the event. Instead, the gross total impact is only 22,500 (= 1.5 x 15,000 = 22,500), and the net total impact is a mere 3,750 (18,750 total from Scenario 1.2 vs. 22,500 total from Scenario 3.2). Furthermore, the marginal effect of the event on the income of local citizens is actually zero since none of the increase in hotel revenues accrues to local residents.

Scenario 4

Finally, suppose the mega-event leaves the number of rooms sold and workers hired constant at 75, but the price of a room doubles to $300. In empirical observations of hotel prices during mega-events, it is not uncommon to observe prices that are double or triple those of non-event room rates. The corresponding direct and indirect earnings from the event for a locally and nationally owned hotel are shown in Table 1, Scenarios 4.1 and 4.2, respectively. As in scenario 3, the increase in room price increases hotel profits while leaving labor's income unchanged. Once more, when capital and labor are both locally supplied, the multiplier is unaffected by the distribution of the proceeds from the event, but when the hotel is not locally owned, the increase in the portion of the hotel's revenue that accrues to capital reduces the multiplier.

Other Considerations

Several other factors can affect the way in which the multiplier changes during a mega-event. First, the presence of local hotel taxes does likely ensure that at least some portion of the hotel's windfall is retained locally. Of course, general sales tax and hotel tax collections are often split between the local municipality and the state, but suppose that it is the local government that imposes a 10% tax on room fees that it will get to keep in its entirety and that the incidence of the tax falls completely on the consumer. Furthermore, assume local tax collections recirculate through the economy at the same rate as other local income so that the multiplier on this tax revenue is 2. Table 1, Scenarios 1.3 through 4.3 show the direct and indirect revenues for the base case and scenarios 1 though 4 for a nationally owned chain. Comparing the multipliers in Scenario 1.2 to Scenario 1.3, Scenario 2.2 to Scenario 2.3, Scenario 3.2 to Scenario 3.3, and Scenario 4.2 to Scenario 4.3, the presence of government taxation serves to raise the multiplier as compared to the situation without taxation.

Second, capital need not be the only source of leakage from a mega-event. If workers from outside the local area come to the host city to provide labor during a mega-event, their earnings are likely to be repatriated back to their home city once the event is over. One particularly notorious but well publicized example of this occurred during the 2006 World Cup in Germany. The country's brothels imported an estimated 40,000 sex workers during the World Cup to accommodate anticipated demand (CBS News, 2006). These foreign workers are unlikely to spend the same portion of their earnings within the German host cities as domestic workers who already resided within the venues. Of course, many other types of labor ranging from specialized security providers to street vendors may temporarily relocate to the host city of a mega-event to provide services.

Going back to Scenario 2 where the mega-event increases the number of rooms sold to 100 while leaving the room prices unchanged, suppose the 25 extra workers the hotel hires in order to accommodate the higher demand are now imported from outside the host city and suppose none of these workers' earnings are spent locally. The corresponding direct and indirect earnings from the event if the hotel is locally owned is shown in Table 1, Scenario 5.1, while the figures for a nationally owned hotel are shown in Table 1, Scenario 5.2. As seen in the table, the fact that the imported workers do not re-spend their earnings in the local economy reduces the multipliers in both cases as compared to the corresponding situations in Scenarios 2.1 and 2.2.

Conclusion and Recommendations

In estimating economic impacts from mega-events, analysts frequently use multipliers derived from input-output tables based on the normal state of the economy even though the presence of a large temporary tourist attraction such as the Olympics or the World Cup indicates a departure from this normal state. Mega-events are characterized by high utilization rates and increased prices for tourism-related industries. While local labor may benefit to some extent through increases in hours worked or higher tips, the main recipient of this windfall is likely to be business owners (and perhaps workers from outside the region). Expenditures in industries dominated by nationally owned chains such as large hotels, rental car agencies, and airlines, and to a lesser extent motels, restaurants, and general retailers may rise significantly due to a mega-event, but local incomes will not increase substantially. Since the benefits accrue to non-local capital owners leading to higher than normal leakages of income, the money generated from these events is unlikely to recirculate through the economy, and any multipliers applied are, therefore, probably inflated.

Are there any obvious solutions to this apparent problem in economic impact analysis? In many ways the situation described in this paper is similar to that described by Lucas (1976) in pointing out the potential problems associated with implementing macroeconomic policy. The so-called "Lucas Critique" suggests that it is naive to presume that one can predict the effects of macroeconomic policy changes based solely on historical data much as this paper shows that it may be equally foolish to presume that one can predict the effects of mega-events on local economies using fixed multipliers based on average economic patterns. Lucas and other rational expectations economists such as Kyland and Prescott advocated the use of computable general equilibrium (CGE) models, and indeed, Dwyer, Forsythe, and Spurr (2004) use this technique in their work on the economic impact of sporting events in Australia. In general, they find that the use of CGE techniques that allow the multipliers to vary along with the state of the economy serve to reduce estimated economic impacts from sporting events in comparison to static models using fixed multipliers.

As few sports boosters seem intent on adopting more conservative approaches to estimating economic impact, caution is warranted whenever economic impact estimates are presented. Cities routinely offer to spend large sums of money in order to attract major sporting events in large part based upon exaggerated claims of an economic windfall, but a skeptical public should beware of economists bearing reports showing "mega-benefits" from mega-events.

References

Baade, R., & Matheson, V. (2001). Home run or wild pitch? Assessing the economic impact of Major League Baseball's All-Star Game. Journal of Sports Economics, 2(4), 307-327.

CBS News. (2006, June 8, 2006). Vatican laments World Cup prostitution.

Crompton, J. (1995). Economic impact analysis of sports facilities and events: Eleven sources of misapplication. Journal of Sport Management, 9(1), 14-35.

Dwyer, L, Forsyth, P., & Spurr, R. (2004). Evaluating tourism's economic effects: New and old approaches. Tourism Management, 25(3), 307-317.

Humphreys, J. (1994). The economic impact of hosting Super Bowl XXVIII on Georgia. Georgia Business and Economic Conditions, May-June, 18-21.

Lucas, R. (1976). Econometric policy evaluation: A critique. Carnegie-Rochester Conference Series on Public Policy, 1, 19-46.

Porter, P., & Fletcher, D. (2008). The economic impact of the Olympic Games: Ex ante predictions and ex poste reality. Journal of Sport Management, 22(4), 470-486.

Siegfried, J., & Zimbalist, A. (2000). The economics of sports facilities and their communities. Journal of Economic Perspectives, 14(3), 95-114.

Siegfried, J., & Zimbalist, A. (2002). Note on the local economic impact of sports expenditures. Journal of Sports Economics, 3(4), 361-366.

Victor A. Matheson (1)

(1) College of Holy Cross

Victor A. Matheson is an associate professor in the Department of Economics. His research interests include the economics of collegiate and professional sports.

Table 1: Economic Impact Estimates and the Resulting Multiplier

                                 Direct Economic Impact

Model/Scenario              Labor   Capital    Gov.    Total

Scenario 1.1: Base case     7,500     3,750      --   11,250
Local ownership

Scenario 1.2: Base case     7,500     3,750      --   11,250
Foreign ownership

Scenario 1.3: Base case     7,500     3,750   1,125   12,375
w/taxation--Foreign
ownership

Scenario 2.1: Increased    10,000     5,000      --   15,000
demand for rooms and
labor--Local ownership

Scenario 2.2: Increased    10,000     5,000      --   15,000
demand for rooms and
labor--Foreign ownership

Scenario 2.3: Increased    10,000     5,000   1,500   16,500
demand for rooms and
labor w/taxation--
Foreign ownership

Scenario 3.1: Increased     7,500     7,500      --   15,000
demand for rooms
only--Local ownership

Scenario 3.2: Increased     7,500     7,500      --   15,000
demand for rooms only--
Foreign ownership

Scenario 3.3: Increased     7,500     7,500   1,500   16,500
demand for rooms only
w/taxation--Foreign
ownership

Scenario 4.1: Increased     7,500    15,000      --   22,500
price of rooms--Local
ownership

Scenario 4.2: Increased     7,500    15,000      --   22,500
price of rooms--Foreign
ownership

Scenario 4.3: Increased     7,500    15,000   2,250   24,750
price of rooms
w/taxation--Foreign
ownership

Scenario 2.1: Increased    10,000     5,000      --   15,000
demand for rooms and
labor--Local ownership,
local labor

Scenario 5.1: Increased    10,000     5,000      --   15,000
demand for rooms and
labor--Local ownership,
foreign labor

Scenario 2.2: Increased    10,000     5,000      --   15,000
demand for rooms and
labor--Foreign
ownership, local labor

Scenario 5.2: Increased    10,000     5,000      --   15,000
demand for rooms and
labor--Foreign ownership
and labor

                                Indirect Economic Impact

Model/Scenario              Labor   Capital    Gov.    Total

Scenario 1.1: Base case     7,500     3,750      --   11,250
Local ownership

Scenario 1.2: Base case     7,500         0      --    7,500
Foreign ownership

Scenario 1.3: Base case     7,500         0   1,125    8,625
w/taxation--Foreign
ownership

Scenario 2.1: Increased    10,000     5,000      --   15,000
demand for rooms and
labor--Local ownership

Scenario 2.2: Increased    10,000         0      --   10,000
demand for rooms and
labor--Foreign ownership

Scenario 2.3: Increased    10,000         0   1,500   11,500
demand for rooms and
labor w/taxation--
Foreign ownership

Scenario 3.1: Increased     7,500     7,500      --   15,000
demand for rooms
only--Local ownership

Scenario 3.2: Increased     7,500         0      --    7,500
demand for rooms only--
Foreign ownership

Scenario 3.3: Increased     7,500         0   1,500    9,000
demand for rooms only
w/taxation--Foreign
ownership

Scenario 4.1: Increased     7,500    15,000      --   22,500
price of rooms--Local
ownership

Scenario 4.2: Increased     7,500         0      --    7,500
price of rooms--Foreign
ownership

Scenario 4.3: Increased     7,500         0   2,250    9,750
price of rooms
w/taxation--Foreign
ownership

Scenario 2.1: Increased    10,000     5,000      --   15,000
demand for rooms and
labor--Local ownership,
local labor

Scenario 5.1: Increased     7,500     5,000      --   12,500
demand for rooms and
labor--Local ownership,
foreign labor

Scenario 2.2: Increased    10,000         0      --   10,000
demand for rooms and
labor--Foreign
ownership, local labor

Scenario 5.2: Increased     7,500         0      --    7,000
demand for rooms and
labor--Foreign ownership
and labor

Model/Scenario             Mult.

Scenario 1.1: Base case     2.00
Local ownership

Scenario 1.2: Base case     1.67
Foreign ownership

Scenario 1.3: Base case     1.70
w/taxation--Foreign
ownership

Scenario 2.1: Increased     2.00
demand for rooms and
labor--Local ownership

Scenario 2.2: Increased     1.67
demand for rooms and
labor--Foreign ownership

Scenario 2.3: Increased     1.70
demand for rooms and
labor w/taxation--
Foreign ownership

Scenario 3.1: Increased     2.00
demand for rooms
only--Local ownership

Scenario 3.2: Increased     1.50
demand for rooms only--
Foreign ownership

Scenario 3.3: Increased     1.55
demand for rooms only
w/taxation--Foreign
ownership

Scenario 4.1: Increased     2.00
price of rooms--Local
ownership

Scenario 4.2: Increased     1.25
price of rooms--Foreign
ownership

Scenario 4.3: Increased     1.39
price of rooms
w/taxation--Foreign
ownership

Scenario 2.1: Increased     2.00
demand for rooms and
labor--Local ownership,
local labor

Scenario 5.1: Increased     1.83
demand for rooms and
labor--Local ownership,
foreign labor

Scenario 2.2: Increased     1.67
demand for rooms and
labor--Foreign
ownership, local labor

Scenario 5.2: Increased     1.50
demand for rooms and
labor--Foreign ownership
and labor