SCAPE Presentation at the SER Conference 4 August 2005 THE MACROECONOMIC EFFECTS OF SINGAPORE S INTEGRATED RESORTS: AN ECONOMETRIC SIMULATION by Tilak Abeysinghe and Choy Keen Meng * Econometric Studies Unit, Department of Economics National University of Singapore How often have I said to you, Sherlock Holmes asked Watson, that when you have eliminated the impossible, whatever remains, however improbable, must be the truth? Now that the Singapore government has decided to proceed with the construction of two Integrated Resorts (IRs), a question often asked is: how much do they help to boost tourism and economic growth? In this study, we provide a partial analysis of the economic impact of the IRs through a simulation of the ESU01 macroeconometric model of Singapore. Specifically, we examine the repercussions for the macroeconomy that might be generated by the development of the resorts. The focus is on tracing out the dynamic paths of key macroeconomic variables such as GDP, investment, service sector output, employment and wages. Even as we do so, we are cognizant of the economic costs associated with law enforcement as well as the wider social costs of gambling. We are not able to quantify these effects, but neither do we dismiss them as trivial. To obtain the baseline solution, we first solve the ESU01 model dynamically for 48 quarterly periods, wherein all exogenous variables are held at their known historical values. We then replace the values of certain exogenous variables in this baseline solution of the model by their counterfactual paths under the alternative IR scenario. Finally, we recompute the solutions for the endogenous variables and compare them with those in the control run to assess the responses of the system. Since ours is a deterministic simulation, the random disturbances in the model stochastic equations are set to zero and only the following exogenous variables are assumed to be affected: * We thank Enrico Tanuwidjaja for research assistance. Roland Cheo and Lim Kim Lian offered valuable suggestions. Peter Wilson contributed comments and the quote from Sherlock Holmes. The documentation for the model is found in a forthcoming book by the authors. 1
As Singapore becomes a more attractive tourist destination, business and leisure travellers rise by 10%. Visitor numbers will also be spurred by the arrival of budget airlines in the region which has led to cheaper airfares. Private construction contracts awarded increases by 20% on account of the two IR projects and spillover expenditures on supporting infrastructure eg. new hotels, restaurants and tourist attractions. Share prices soar by 10% as the valuation of travel-related and property stocks appreciate sharply. This indirectly encourages private consumption spending via an increase in financial wealth. Government consumption spending is raised by 1% in order to set up a new regulatory agency for the casinos, bolster law enforcement efforts and provide social services and safeguards. The extra spending is assumed to be financed by tax revenues from the operation of the IRs. In reality, of course, these events need not occur simultaneously. If we imagine for simulation purposes that they all happen together at time zero, however, their macreconomic effects can be visualized in Figures 1 and 2. These charts show the percentage deviations in the simulation run of the key macroeconomic variables from their baselines values at subsequent time periods. The first result to note is that the combined output of the service industries shoots up by 0.13% immediately and 0.28% after half a year (this is mirrored by an instantaneous increase in service exports of 2.5%, not shown). In terms of the sectoral impact in Figure 2, the industries that benefit most are those in commerce hotels and restaurants and wholesale and retail trade. The value-added of the commerce sector rises by nearly half a percent compared to the baseline scenario. By contrast, the transport and communication sector s output increases by a smaller 0.2%, though for a more protracted period. The financial and business sector also stands to benefit as the final demand for such services rises in tandem with the economic boom. Turning to capital spending, construction investment rises right after the commencement of the building projects by about 0.23%. Due to long gestation lags, the initial increase in construction contracts leads to larger effects subsequently on construction gross fixed capital formation (and construction value-added in Figure 2), amounting to about 0.6% at their maximum. Interestingly, machinery and transport equipment investment is also stimulated by the growth in service sector output, suggesting that a bigger influx of tourists creates a virtuous investment cycle. Concurrently, private consumption expenditures rise by 0.8% relative to their baseline path as a consequence of the increase in the disposable incomes and financial wealth of households. 2
Figure 1: Dynamic Effects on Selected Macroeconomic Variables GDP.30 SERVICE.16.25.15.05 - -.05.012 CPI 1.0 CONSUMPTION 8 0.8 4 0.6 0.4 0 0.2-4 0.0-8 -0.2 EQUIPMENT INVESTMENT.32.28.24.16 - CONSTRUCTION INVESTMENT.7.6.5.4.3.2.1.0 EMPLOYMENT WAGES.06.06.02.02 -.02 -.02 3
Figure 2: Dynamic Effects on Sectoral Value-Added.6.5.4 COMMERCE CONSTRUCTION FINANCIAL AND BANKING SERVICES OTHERS TRANSPORT AND COMMUNICATIONS.3.2.1.0 -.1 Like services output, the peak in overall economic activity as measured by real GDP is reached after 2 quarters. At 0.2% above baseline, this effect might seem small. However, the cumulative impact on aggregate output is not so, as we shall see below. Higher national income in turn generates positive multiplier effects on employment and wages, albeit with some delay. The hump-shaped responses of these labour market variables reflect a process of slow adjustment taking place. Similarly, the inflationary effects of the rise in output on the CPI kick in only after two years, but they tend to be small anyway. Figure 3 depicts the cumulative effects of the IRs on GDP, services output and total employment, obtained by summing the dynamic effects plotted in Figure 1. The model simulation results indicate that Singapore s economy will be lifted for about five years. In a year s time for example, the level of real GDP would have risen to 0.75% above the baseline and the production of services would be 1% higher. These figures balloon to 1.6% and 2% after three years, and to 1.8% and 2.2% in the long term. Employment will expand correspondingly by 0.9%, translating into roughly 20,000 new jobs. Under our assumptions, the increase in total output is comparable in magnitude, but opposite in direction, to those of the SARS outbreak in 2003. In conclusion, we stress that the scenario we painted, with its attendant implications, is by no means the most comprehensive one. For one thing, we have neglected the direct employment creation resulting from the resorts, which has been estimated to be in excess of 10,000 workers. There could also be other effects, both positive and negative, which we have not incorporated in the simulation. See the earlier study by the authors entitled Revisiting SARS: Impact on the Singapore Ecconomy, available at http://nt2.fas.nus.edu.sg/ecs/cent/esu/conference.htm. 4
Figure 3: Cumulative Impact on Output and Employment 2.5 2.0 1.5 SERVICE GDP 1.0 EMPLOYMENT 0.5 0.0 5