In the previous sections, we showed that using a constant JSR to predict employment effects of additional demand for tourism services would have led to overestimation in high capacity utilization situations and underestimation when demand levels are low. This pattern is observed for the Taiwanese hotel and aviation industries. We think that the level of estimation error is directly related to 1) the magnitude of demand changes, 2) the duration of the event, 3) the size of geographic region, and 4) the industry characteristics. A larger estimation error is expected for tourism applications involving dramatic demand fluctuations, within a short period for a small region. In terms of industry differences, sectors that can adjust their total service capacities, employ high-skilled personnel, and have less part-time or temporary staff would then yield larger job estimation errors. The aviation sector, the land and sea transportation sectors would fall under this category. On the other hand, the accommodation sector and the amusement sector have no flexibility in adjusting their service capacity due to physical infrastructure constraints, and most of their job positions are open for low-skilled, part time or temporary staff. For these sectors, a smaller range of variation for their jobs-to-sales ratios is expected.
To increase the accuracy of job estimates, the price effect and the labor efficiency effect should be taken into consideration simultaneously. For the price effect, some scholars have attempted to correct for the price bias by introducing the consumer price index (CPI) into the model to adjust for intertemporal price differences(Stynes et al., 2000). For short-term IO applications, this attempt only partially corrects for bias, because the short-term price change in the tourism industry is generally more substantial than suggested by the yearly CPI. For example, Porter (1999) indicated that the real room price during the 1989 and 1995 Miami Super Bowls and the 1991 Tampa Super Bowl rose by 11.26%, 19.83% and 4.44%, respectively, during the one-month period from the previous year. Practitioners therefore are recommended to take real price changes into consideration as the level of final price adjustment will be more prominent than the yearly data from the government statistics.
We also advise practitioners to include assessments of labor efficiency in the project evaluation process. Our Scenario 2 in the previous section demonstrates that changes in labor efficiency are the primary factor contributing to JSR variation. Total annual earnings, total number of persons engaged, and capacity utilization are generally made available through government statistics as most countries have routine business censuses for the manufacturing and service industries. Econometric models can be estimated using such official data.
Last, the rigidity of the recruiting system has a close link to the stages in the destination life cycle. If the region is in the exploration, involvement or development stage, infrastructures and new businesses are added to the region, and per unit sales can generally support more jobs than other stages. On the other hand, if the region has set up the lodging, dining and recreational capacity or becomes matures, additional consumption will not significantly change the numbers employed unless the opening of new business is observed.
Similarly, during a tourism downtime, business entities will tend to keep their staff unless the operations of business have to be curtailed. Taiwan tourist hotels are currently situated at the
period. This is the reason why room capacity is insignificant in influencing the jobs-to-sales ratio. Since most studies focus on providing a quantitative number in economic impacts, practitioners should also collect data on the opening and closure of business entities during the evaluation period. This is where the most meaningful job creation and job losses take place.
Finally, we should stress that we only considered the direct effects of changing conditions for the tourism industry. More activity in the hotel and/or aviation sectors, for example, will also lead to increased demand for intermediate inputs in the production processes of these sectors. Up to some extent, the assumption of constant JSRs in these sectors might also be an unrealistic assumption. Furthermore, some input-output models assume that consumption levels should also be endogenized, reflecting the idea that higher income will lead to more consumption. Our study suggests that the number of new employees is not changing quickly in response to changes in tourism demand and it remains to be seen whether already employed staff members receive much more wages. If not, consumption levels will not increase very much and consequent effects of employment will not materialize.
6. Conclusion
An increasing number of economic impact studies are carried out to address special tourism demand conditions, such as hosting mega events or facing extreme weather, disease outbreaks or terrorist activities. Under these scenarios, productivity levels and cost structures of the tourism industry undergo substantial changes. Using a long-term IO model to predict the consequences of short-term events inevitably lead to estimation biases. This paper clarifies the underlying relationships between the jobs-to-sales ratio and capacity utilization.
The results indicate that the adjustment of labor efficiency is the prominent factor in determining the stability of the jobs-to-sales ratio, while price, total employee number and service capacity are relatively stable in response to demand changes. For the hotel sector, a 10% fluctuation in the occupancy rate leads to a 15-20% adjustment in the jobs-to-sales ratio, 2% in room price, 2% in employees and 13-18% in labor efficiency. For the aviation sector, a 10% fluctuation in the load factor resulted in a 24-38% adjustment in the jobs-to-sales ratio, 5% in total passenger miles offered, and 21-32% in labor efficiency. The magnitude of biases associated with assuming a stable job-to-sales ratio is too large to be ignored. While Frechtling (2006), Stynes & White (2006) and Wilton & Nickerson (2006) indicated that the possibly largest errors of tourism economic impact estimation are caused by the inaccuracy of forecasts (or measurement) of visitor numbers and average spending, we argue that stability of economic and technological ratios representing production processes in the tourism industry should also be considered with care.
The implications of this study can also be integrated into Computable General Equilibrium models (CGE), which generally assume constant returns to scale and incorporate the price mechanism to clear all markets simultaneously (Blake & Gillham, 2001; Dwyer et al., 2004). This study indicates that only taking the price adjustment mechanism into account is insufficient in reaching a satisfactory portrayal of the reality in labor usages per unit output.
The adjustment of labor efficiency during various level of capacity utilization should be acknowledged and endogenized in the system to allow for changing returns to utilization so that more accurate job estimates can be provided.
Acknowledgement
Constructive comments from Dr. Bart Los at University of Groningen, and two anonymous referees are highly appreciated. Financial support from the Taiwan National Science Council under NSC98-2410-H-390-029-SS2 is acknowledged.
7. References
Ahlert, G. (2001). The Economic Effects of the Soccer World Cup 2006 in Germany with Regard to Different Financing. Economic Systems Research, 13, 109 - 127.
Archer, B. (1977). Tourism multipliers: The state of art. Cardiff: University of Wales Press.
Arenberg, Y. (1991). Service capacity, demand fluctuations, and economic behavior.
American Economist, 35, 52-55.
Baade, R. A., & Matheson, V. A. (2003). Bidding for the Olympics: Fool's gold? In C.
Barros, M. Ibrahim & S. Szymanski (Eds.), Transatlantic Sport (pp. 127-151).
London: Edward Elgar Publishing.
Baade, R. A., & Matheson, V. A. (2004). The quest for the cup: Assessing the economic impact of the World Cup. Regional Studies, 38, 343-354.
Baade, R. A., & Matheson, V. A. (2006). Padding required: Assessing the economic impact of the super bowl. European Sports Management Quarterly, 6, 353-374.
Baltagi, B. H., Griffin, J. M., & Vadali, S. R. (1998). Excess capacity: A permanent characteristic of US airlines? Journal of Applied Econometrics, 13, 645-757.
Berndt, E. R., & Morrison, C. J. (1981). Capacity utilization measures: Underlying economic theory and an alternative approach. American Economic Review, 71, 48-52.
Blake, A., & Gillham, J. (2001). A Multi-Regional CGE Model of Tourism in Spain. Paper presented at the European Trade Study Group annual conference, Brussels.
Blake, A., & Sinclair, M. T. (2003). Tourism crisis management: US Response to September 11. Annals of Tourism Research, 30, 813-832.
Blake, A., Sinclair, M. T., & Sugiyarto, G. (2003). Quantifying the impact of foot and mouth disease on tourism and the UK economy Tourism Economics, 9, 449-465.
Borooah, V. K. (1999). The supply of hotel rooms in Queensland, Australia. Annals of Tourism Research, 26, 985-1003.
Bryden, J. M. (1973). Tourism and development: A case study of the Commonwealth Caribbean. Cambridge, England: University Press.
Caves, D. W., & Christensen, L. R. (1988). The importance of economies of scale, capacity utilization, and density in explaining interindustry differences in productivity growth.
Logistics and Transportation Review, 24, 3-32.
Coates, D., & Humphreys, B. R. (2003). Professional sports facilities, franchises and urban economic development. Public Finance and Management, 3, 335-357.
Conway, R. S. (1980). Changes in regional Input-Output coefficients and regional forecasting. Regional Science and Urban Economics, 10, 153-171.
Crompton, J. L. (1995). Economic impact analysis of sports facilities and events: Eleven sources of misapplication. Journal of Sport Management, 9, 14-35.
Dwyer, L., Forsyth, P., Spurr, R., & Ho, T. (2004). The economic impacts and benefits of tourism in Australia: A general equilibrium approach. In C. Cooper (Ed.). Southport, AU: Cooperative Research Centre for Sustainable Tourism.
Fletcher, J., & Snee, H. (1989). Tourism multiplier effects. In S. F. Witt & L. Moutinho (Eds.), Tourism Marketing and Management Handbook (pp. 529-531). Hertfordshire, UK: Prentice Hall International Ltd. (UK).
Forssell, O. (1972). Explaining changes in input-output coefficients for Finland. In A. Brody
& A. P. Carter (Eds.), Input-Output Techniques (pp. 343-369). Amsterdam:
North-Frechtling, D. C. (2006). An assessment of visitor expenditure methods and models. Journal of Travel Research, 45, 26-35.
Gasparino, U., Bellini, E., Corpo, B. D., & Malizia, W. (2008). Measuring the impact of tourism upon urban economies: A review of literature FEEM working paper series.
Milan: The Fondazione Eni Enrico Mattei.
Gujarati, D. N., & Porter, D. C. (2009). Basic Econometrics (5th ed.). New York: The McGraw-Hill companies, Inc.
Hagn, F., & Maenning, W. (2009). Large sport events and unemployment: The case of the 2006 soccer World Cup in Germany. Applied Economics, 41, 3295-3302.
Hatch, J. H. (1986). The impact of the Grand Prix on the restaurant industry. In J. P. A.
Burns, J. H. Hatch & T. J. Mules (Eds.), The Adelaide Grand Prix - The impact of a special event: The Centre for South Australian Economic Studies.
Krakover, S. (2000). Partitioning seasonal employment in the hospitality industry. Tourism Management, 21, 461-471.
Lee, C.-K., & Taylor, T. (2005). Critical reflections on the economic impact assessment of a mega-event: the case of 2002 FIFA World Cup. Tourism Management, 26, 595-603.
Miller, R. E., & Blair, P. D. (1985). Input-output analysis: Foundations and extensions.
Englewood Cliffs, N.J.: Prentice-Hall.
Porter, P. K. (1999). Mega-sports events as municipal investments: A critique of impact analysis. In J. Fizel, E. Gustafson & L. Hadley (Eds.), Sports Economics: Current Research (pp. 61-74). Westport, Connecticut: Praeger Publishers.
Porter, P. K., & Fletcher, D. (2008). The economic impact of the Olympic Games: Ex ante predictions and ex poste reality. Journal of Sport Management, 22, 470-486.
Price Water House Coopers. (2005). Olympic Games Impact Study. London: Price Water House Coopers LLP.
Rose, A., & Miernyk, W. (1989). Input-output analysis: The first fifty years. Economic Systems Research, 1, 229-271.
Stynes, D. J., Propst, D. B., Chang, W., & Sun, Y. (2000). Estimating national park visitor spending and economic impacts: The MGM2 model. Final report to National Park Service. East Lansing, Michigan: Department of Park, Recreation and Tourism Resources, Michigan State University.
Stynes, D. J., & White, E. M. (2006). Reflections on measuring recreation and travel spending. Journal of Travel Research, 45, 8-16.
Sun, Y.-Y. (2007). Adjusting Input-Output models for capacity utilization in service industries. Tourism Management, 28, 1507-1517.
Szymanski, S. (2002). The economic impact of the World Cup. World Economics, 3, 169-177.
Taiwan Stock Exchange Cooperation. (2009). Market Observation Post System (MOPS), from http://emops.tse.com.tw/emops_all.htm
Taiwan Tourism Bureau. (2002-2009). The operating report of international tourist hotels in Taiwan. Taipei, Taiwan: Tourism Bureau, Ministry of Transportation and Communications, R.O.C.
Toohey, K., & Veal, A. J. (2007). The Olympic Games: A social science perspective (2nd ed.). Oxfordshire, UK: CABI International.
Vaccara, B. N. (1971). An Input-Output model for long-range economic projections. Survey of Current Business, 51, 47-56.
West, G. (1995). Comparison of Input-Output, Input-Output Econometric and Computable General Equilibrium Impact Models at the regional level. Economic Systems Research, 7, 209-227.
West, G., & Gamage, A. (2001). Macro effects of tourism in Victoria, Australia: A nonlinear
Wilton, J. J., & Nickerson, N. P. (2006). Collecting and using visitor spending data. Journal of Travel Research, 45, 17-25.
Wooldridge, J. M. (2009). Introductory Econometrics: A Modern Approach (4th ed.).
Cincinnati, OH: South-Western.
World Health Organization. (2003). Taiwan, China: SARS transmission interrupted in last outbreak area Retrieved June 16, 2007,
from http://www.who.int/csr/don/2003_07_05/en/
Yang, H.-Y., & Chen, K.-H. (2009). A general equilibrium analysis of the economic impact of a tourism crisis: A case study of the SARS epidemic in Taiwan. Journal of Policy Research in Tourism, Leisure and Events, 1, 37-60.
Table 1. Descriptive statistics of monthly tourist hotel operations and quarterly airline operation in Taiwan (2002-2009)
Monthly data Min Max Mean
Std.
Deviation
Coefficie nt of variation (CV) Hotel
Total rooms available 19936 22311 20,941 480 2.3%
Occupied rooms / month 136,814 485,271 413,540 54,650 13.2%
Occupancy rate 22% 78% 65% 8% 12.6%
Average room price (NT$) 2,301 3,438 3,014 195 6.5%
Total employees 6,013 7,912 7,265 311 4.3%
Jobs to monthly million sales ratio 4.81 18.62 6.04 1.70 28.1%
Airlines
Total number of flights per quarter 13,873 25,253 21,817 2,842 13.0%
Sold passenger miles per quarter (millions)
6,087 15,208 12,973 1,768 13.6%
Available passenger miles per quarter (millions)
12,132 18,764 16,905 1,544 9.1%
Passenger load factor 50% 81% 76% 0.06 7.4%
Average price per thousand passenger miles (NT$)
1.03 2.30 1.78 0.28 15.7%
Total employees 13,490 15,309 14,412 588 4.1%
Jobs to monthly million passenger
miles sales ratio 0.45 1.58 0.67 0.22 32.9%
Table 2. Statistics on tourist hotels before, during, and after the SARS period, 2003 Month Occupancy
rate
Average room price (NT$)
Total room sales (NT$
millions)
Number of tourist
hotels
Hotel employees
Jobs to sales ratio (jobs per million NT$)
Feb 68% $2,894 $1,145 83 7,474 6.53
Mar 61% $2,825 $1,114 82 7,222 6.48
Apr (SARS) 37% $2,653 $614 81 6,956 11.33
May (SARS) 22% $2,459 $336 78 6,267 18.62
Jun (SARS) 34% $2,301 $464 77 6,013 12.97
Jul 58% $2,781 $1,031 80 6,533 6.34
Aug 68% $2,759 $1,202 81 6,750 5.61
Table 3. Variance components analysis of lnJSR and lnCU for the hotel sector
Var (lnJSR) = VAR(lnJ) +Var(lnX) +Var(lnP) –2*Cov(lnJ, lnX)
–2*Cov(lnJ, lnP)
+2*Cov(lnX , lnP) 0.0355 = 0.0020 0.0296 0.0044 0.0083 0.0039 0.0117
100% = 3% 50% 7% 14% 6% 20%
Var (lnCU) = Var(lnX) + Var(lnT) -2*Cov(lnX, lnT) 0.0267 = 0.0296 0.0014 0.0044
100% = 84% 4% 12%
Cov[lnJSR, lnCU] = Cov[(lnJ -lnX –lnP), (lnX – lnT)]
= cov(lnJ,
lnX) – var(lnX) – cov(lnP, lnX)
– cov(lnJ, lnT)
+ cov(lnX, lnT)
+ cov(lnP, lnT) -0.0292 = 0.0041 0.0296 0.0058 0.0007 0.0022 0.0006
100% = 10% 69% 14% 2% 5% 1%
Table 4. Variance components analysis of lnJSR and lnCU for the airline sector
Var (lnJSR) = VAR(lnJ) +Var(lnX) +Var(lnP) –2*Cov(lnJ, lnX)
–2*Cov(lnJ, lnP)
+2*Cov(lnX , lnP) 0.0672 = 0.0017 0.0284 0.0288 0.0092 0.0079 0.0254
100% = 2% 28% 28% 9% 8% 25%
Var (lnCU) = Var(lnX) + Var(lnT) -2*Cov(lnX, lnT) 0.0075= 0.0284 0.0095 0.0303
100% = 42% 14% 45%
Cov[lnJSR, lnCU] = Cov[(lnJ -lnX –lnP), (lnX – lnT)]
= cov(lnJ,
lnX) – var(lnX) – cov(lnP, lnX)
– cov(lnJ, lnT)
+ cov(lnX, lnT)
+ cov(lnP, lnT) -0.0175= 0.0046 0.0284 0.0127 0.0034 0.0152 0.0072
100% = 6% 40% 18% 5% 21% 10%
Table 5. Regression results for the hotel data, 2002-2009
Variable Log jobs to
sales ratio Log of room
price Log of
Durbin h test p-value 0.632 Breusch-Pagan test
p-value 0.162 0.605 0.611 0.601
Cook-Weisberg test p-value
0.085 0.086 0.405 0.818
Standard error statistics are in parentheses.**Significant at the 99% level; *significant at the 95% level.
Table 6. Regression results for the airline data, 2002-2009
Variable Log jobs to
sales ratio Log of mile
price Log of
employee Log of total mile capacity
intercept -0.788** -6.357** 9.522** 15.313**
(0.051) (0.050) (0.013) (2.797)
lnORt -1.768** 0.231 -0.009 0.633**
(0.152) (0.149) (0.038) (0.071)
lnYt-1 0.355**
(0.120)
year 2003 -0.020 -0.018 -0.005 0.041
(0.045) (0.045) (0.011) (0.021)
year 2004 -0.078 0.031 0.028* 0.030
(0.042) (0.042) (0.011) (0.022)
year 2005 -0.184** 0.086 0.066** 0.092**
(0.042) (0.042) (0.011) (0.027)
year 2006 -0.238** 0.152** 0.100** 0.096**
(0.042) (0.042) (0.011) (0.031)
year 2007 -0.286** 0.198** 0.093** 0.089*
(0.043) (0.042) (0.011) (0.032)
year 2008 -0.353** 0.284** 0.073** 0.059
(0.042) (0.042) (0.011) (0.029)
year 2009 0.243** -0.358* 0.046** 0.109**
(0.052) (0.051) (0.013) (0.026)
F 65.73** 27.58** 24.22** 42,81**
Adj. R2 0.947 0.880 0.865 0.931
Durbin h test p-value 0.837
Breusch-Pagan test
p-value 0.348 0.115 0.477 0.838
Cook-Weisberg test p-value
0.500 0.066 0.747 0.794
Standard error statistics are in parentheses. **Significant at the 99% level; *significant at the 95% level.
Table 7. Predicted results based on different capacity utilization rates
Capacity utilization rate 50% 60% 70%
Pct change
Jobs to sales ratio 8.19 6.80 5.81 20.5% -14.6%
price per room $2,844 $2,900 $2,948 -1.9% 1.7%
jobs 7,746 7,893 8,019 -1.9% 1.6%
capacity 20,715 20,818 20,906 -0.5% 0.4%
jobs/sold units 0.062 0.053 0.046 18.4% -13.3%
Airline
Jobs to sales ratio 1.55 1.12 0.85 38.0% -23.9%
price per 1,000 mile $1.03 $1.08 $1.12 -4.1% 3.6%
jobs 13,739 13,717 13,699 0.2% -0.1%
capacity 28,059 29,893 31,846 -6.1% 6.5%
jobs/ million sold passenger miles
2.04 1.54 1.22 32.3% -21.1%
Note: The J/X ratio is computed by dividing the predicted jobs by the predicted occupied rooms at the given occupancy rate.
Table 8. Estimation results based on two scenarios
Occupancy rate 50% 60% 70%
Pct
Scenario 1: Ten million dollars of final demand changes
direct jobs - fixed ratio1 68.0 68.0 68.0 0.0% 0.0%
direct jobs - predicted ratio2 81.9 68.0 58.1 20.5% -14.6%
Scenario 2: Per 1,000 occupied rooms
Room sales (millions) $2.84 $2.90 $2.95 -1.9% 1.7%
direct jobs - fixed ratio1 19.3 19.7 20.0 -1.9% 1.7%
direct jobs - predicted ratio2 23.3 19.7 17.1 18.1% -13.2%
1 Standard IO analysis assumes a fixed jobs to sales ratio.
2 The predicted jobs to sales ratio is taken from Table 5.