The Impact of Tourist Spending in South Africa: Spatial Implications

In South Africa tourism is regarded as a growth catalyst and a contributor towards the economic upliftment of poorer regions. The magnitude of the impact of tourist spending depends on the extent of leakages from the region, and the ability of the region to create backward and forward linkages. It is argued that the spatial implications of tourism might, inter alia, depend on the relative impacts of domestic versus international tourist spending. Using an input-output model, support was found for the notion that, from a regional development perspective, there should be an emphasis on the development of the domestic tourism market in the short tenn. As a long-tenn strategy, investment in transport services and infrastructure to reduce path dependency effects, more inherent in international tourism, is proposed. JEL L 84, R 11


INTRODUCTION
Tourism, as well as infonnation technology and telecommunications, are recognised as industries with significant growth and development potential in a global economy (Kruger-Cloete, 1995: 751). The tourism industry in particular has been described as the largest global industry (SATOUR circular, 1996; Van der Merwe, 2000: I). It is also one of the most rapidly expanding industries with tourist arrivals growing by 2.4 % worldwide in 1998 (WTO, 1998;Brynard, 1995: 12;Hicks, 1996: 7). After petroleum and petroleum products and motor vehicles, international tourism is the third largest export industry globally (Kruger-Cloete, 1995: 271).
The tourism industry consists of a number of different but interlinked service industries, particularly accommodation and catering, food processing and beverages, transport, as well as entertainment and other support services (Paton, 1985: 64). Increasingly, many developing countries are recognising a potential comparative advantage in tourism and are, as a consequence, adopting tourism development strategies to further growth and development. Particularly in Africa, tourism has been argued to have significant development potential. Due to its geographic and climatic diversity, ecology, biodiversity and exotic fauna and flora, Africa may be argued to have a comparative advantage in tourism that has so far been unexploited.
Assuming that political and social stability will be maintained, the impact of a growing tourism industry in developing countries may depend on its forward and backward domestic linkages, the nature of its regulation (affecting sustainability) as well as the relative emphasis on domestic versus international tourism. In most countries the development of tourism will also have a specific regional (or spatial) impact, due to the geographic distribution of tourism attractions and supporting services. Relatively little research has so far been focused on this aspect of tourism promotion as a development strategy. This paper is an initial (and tentative) attempt to contribute towards filling this vacuum. It will do so by arguing that the spatial outcome of tourism development may depend, in part, on the relative emphasis on domestic versus international tourism. An input-output model of South Africa, an upper middleincome developing country with a possible unexploited comparative advantage in tourism, will be used to find an empirical basis for the assertion that domestic, relative to international, tourism has a differential spatial impact on economic development.
The paper is structured as follows: Section two discusses the merits and demerits of foreign tourism promotion in South Africa. Section 3 explains the methodology used to determine the contribution of domestic and international tourist spending to the South African economy. In section 4 the results of a 10% increase in domestic and international tourism, respectively, are presented. Section 5 discusses the implications of these results for spatial (or regional) development in South Africa. Section 6 provides a concluding summary of the findings.

2
FOREIGN VERSUS DOMESTIC TOURISM PROMOTION

Definitions
Tourism per se can be defined as "the activity of persons travelling to and staying in places outside their usual environment for not more than one consecutive year for leisure, business or any other purpose" (WTO, 1999). Anyone who spends at least one night away from home is thus classified as a tourist. When travel takes place within the country of residence, it is referred to as domestic tourism (WTO, 1999). While international tourism usually refers to all tourists from foreign countries, this paper uses the term international tourism for tourists from foreign countries, excluding tourists from Africa. Foreign tourism is the term used in the context of this paper to describe the sum of international tourists (as defined above), including tourists from Africa.

The Case for Domestic Tourism
Since 1994, following South Africa's democratic transition and the signing of the Uruguay Round of the GATT, foreign tourism to South Africa has increased substantially (Msimang, 1995: 20). Foreign tourist arrivals have, for instance, shown a steady increase of more than 10% per annum on average since 1988 (Saayman, 1999). This makes South Africa one of the world's fastest-growing tourism destinations and has raised expectations as to the impact of foreign tourism on the development of the economy (Van der Merwe, 2000: 2; Wood, 1995: 29).
This high rate of current growth could be reflecting a "catching-up" phase during which South Africa's tourism share is converging on that of similarly endowed international tourism destinations. For example, South Africa's tourism sector is currently contributing around 4% to GDP, compared to the above-mentioned 10% contribution of the industry to the economies of the United States and Europe (De Beer et aI., 1998: 2).
In a globalising economy, where increased international competition is a defining feature, and path-dependence effects in tourism may have to be overcome for South Africa to be competitive enough to maintain these growth rates of tourist arrivals and thus successfully converge on international trends, government promotion of foreign tourism might be justified. In effect, the argument for government promotion of tourism in South Africa can be based on overcoming the path-dependency effects (due to history), as well as addressing the high fixed costs in tourism, which may create obstacles to entry into international markets for South African tourism service providers. Given, however, that the opportunity cost of government expenditure is high in this country, which is in need of basic social services such as housing and education (Muller, 1997: 10), the impact of international tourism on the domestic economy, for example as measured by its multiplier effect, must be sufficiently high for social welfare to be maximised.
Another argument in favour of significant government promotion of foreign tourism has recently been voiced. This states, simplistically, that foreign tourism is often promoted to the detriment of domestic tourism and that this bias is not justified. In South Africa, domestic tourism contributes approximately 60% to total tourism income (Wood, 1995: 29). It has been argued that the domestic tourism market has significant growth potential, also based on a "catching-up" effect, if it is made accessible to South Africa's previously disadvantaged population groups (Futter & Wood, 1997: 58;Wood, 1995: 29). The expectation is therefore that tourism may be one sector through which the concept of "growth through redistribution" might be applied. It has also been argued that the domestic tourism sector has to act as a buffer against unpredictable fluctuations in international tourism demand, and thus fluctuations in foreign exchange earnings (De Beer et at., 1998: I).

The model
Tourism ( The typical approach in economic literature to quantify the impact of tourism spending on the economy, is to use a standard input-output table and model the effect of an exogenous change in spending on final demand. This analysis makes use of matrix algebra to determine final net output lmd can be used to estimate the level of income, employment and production required to satisfy a cenain level of tourism demand. This analysis also generates multipliers, which include secondary impacts, in an impact analysis (see Kottke, 1988). Alternative methods used elsewhere, such as CGE models and SAM-based mUltiplier models (see e.g. Zhou et at., 1997; Wagner, 1997 or Wang, 1995) were not considered in the present case since the most recent SAM available for South Africa is for 1995 (and is basically an adjustment from the 1992 base year which, in tum, is based on a 1986 input-output table).
The weaknesses of an input-output analysis include that industry as a whole and not individual firms are dealt with, that multipliers are subject to misinterpretation (Kottke, 1988: 123) and that tourism data are usually inadequate for the detailed requirements of an input-output analysis (Archer, 1997). Nevertheless, Wanhill (1994) indicates that it is still the most popular tool to determine the impact of tourism on the economy and has been used, among others, by researchers such as Wang (1997), Andrew (1997), Bergstrom et al. (1990), Fletcher (1989), Heng and Low (1990), Loomis (1995) and Smeral (1999).

The Data
In the present paper, the recently released 1996 South African Input-Output (1-0)

The Simulations
Given that tourists to South Africa have been increasing by 10% per annum on average over the past five years, the simulations attempted to analyse the impact of a 10% increase in both domestic and foreign tourism spending. This of course assumes a linearly homogenous relationship between tourist arrivals and tourist spending.
Whilst overall magnitudes of foreign and domestic tourist spending are available, it is not obvious how these should be allocated in an input-output model to the various economic sectors. Additional information on tourist spending patterns, taken from SATOUR surveys, had to be used for this purpose. These allocations determined the simulations (or shocks) performed.

J Modelling a J 0 per cent increase in domestic tourist expenditure
Based on the 1996 SA TOUR survey, domestic tourist spending contributed RI9 112 billion to GDP. South Africans undertook 22644 million trips in that year, spending an average of R844 per trip (SATOUR, 1997: 17). The breakdown of domestic tourist spending into different categories is indicated in Table 1. These expenditure categories were attributed to various sectors in the 1996 1-0 table according to the proportions of consumer expenditure reflected in the 1-0 table.

Modelling a 10 per cent increase in international tourist expenditure
The volume and expenditure patterns of international tourists to South Africa differ from that of domestic tourists. The likely (or possible) differential impact which it might have on the local economy (and its regional/spatial impact), can be modelled and compared to that of domestic tourist expenditure.
International tourist spending is estimated to have contributed R12.7546 billion to South Africa's GDP in 1996. Approximately 1.2 million international tourists (excluding African tourists) visited South Africa that year (Rhodes & Saayman, 1998). The average spending per tourist was RI8 600 (SA TOUR, 1996: 17). According to the August 1996 survey done by SATOUR, the R 18 600 was spent as follows (Table 2): According to the South African Reserve Bank Quarterly Bulletin (1998), the airfare to travel to South Africa (via South African Airways), amounted to R2 235 million. Using the tourism traffic figure from the SARB quarterly bulletin, the amount of prepaid expenses entering South Africa was determined at R2 707.6 million. If it is assumed that most prepaid expenses are for pre-booked tours, the following assumptions (Table 3) regarding prepaid expenses entering South Africa can be made: Table 3 Assumption regarding prepaid expenditure F or the purpose of this analysis, the international tourist spending categories and totals are summarised in Table 4: These expenditure categories were allocated to the corresponding sectors in the 1996 1-0 table. In total, the above implies that a 10% increase in foreign tourist expenditure increases domestic demand with R1.23 billion.

Impact of Domestic Tourism
A 10% increase in total domestic tourist expenditure increases domestic demand with R 1.82 billion. In order to determine the multiplier effect thereof, and the sectoral impact, the X-vector of final demands in the 1-0 model was adjusted by the above increase. The results can be classified as macroeconomic effects and sectoral effects.
The macroeconomic effect of a 10% increase in domestic tourist expenditure was found to lead to an increase in output of R3.S8 billion. This implies a multiplier coefficient of 1.96. This multiplier indicates the increase in production or output, and such output multipliers are generally greater than income multipliers (Wang, 1997: 42).
The sectoral effects were as follows:  Table S shows that the sector to benefit most from the increase in tourist expenditure is manufacturing (R 1.31 bn), followed by transport and communications (RO.61 bn). It is noteworthy that the catering and accommodation category benefited third most.
The detailed results indicated that within manufacturing, the food and food processing sector (RO.67 bn) and the beverage sector (RO.17 bn) benefited the most.

Impact of International Tourism
As in the case of domestic tourism, the results can be categorised as macroeconomic and sectoral results.
The macroeconomic results indicated an increase in total output of R2.34 billion. Given the original exogenous injection of R 1.23 billion, this implies a multiplier of 1.9 -somewhat lower than that of domestic tourism, but not significantly so.
The sectoral results are given in Table 6:  Table 6 shows that, as in the case of domestic tourists, the manufacturing sector benefited most (RO.63 bn). Within manufacturing, food and food processing is the sector that benefits most. Nex' to manufacturing, the services sector (which includes business services related to recreation) benefited 'the most (RO.53 bn), followed by transport (ROA7 bn) and catering and accommodation (RO.32 bn).
The table also indicates that, compared to domestic tourism, agriculture benefited less from international than domestic tourists (RO.14 bn versus RO.31 bn).

SPATIAL IMPLICATIONS
The above results may have the following spatial (provincial) implications for tourism in South Africa: Firstly, both international and domestic tourism have a posItive impact on economic development in South Africa, with multipliers of 1.9 and higher. An important finding is that the overall multiplier effect of domestic tourism is not significantly different from that of international tourism. According to the White Paper on Tourism (South Africa, 1996) provincial authorities (i.e. regional governments) are responsible for the development and promotion of domestic tourism. The national agency, SATOUR, is responsible for marketing South Africa as an international tourist destination. To maximise the benefits from tourism, provincial governments should become more active in marketing their products to new markets within South Africa.
A second implication is that international tourism has more benefits for services, which are currently more developed in affluent regions, such as the provinces of Gauteng, Western Cape and KwaZulu-Natal. These provinces also attract the most domestic as well as international tourists (see Appendix A) (Rhodes & Saayman, 1997). A further advantage of these three provinces is that they offer the three international gateways to South Africa, increasing their accessibility to tourists. Table 7 shows the percentage share of total gross geographical product (GGP) for different sectors of the economy which can be attributed to different provinces in South Africa. The significance of this is that it supports the argument that tourist spending benefits certain provinces more than others. For example, if trade and accommodation increase because of an increase in demand from more tourists, Gauteng will benefit most, followed by Western Cape and KwaZulu-Natal. It is the same with services, transport, printing and publishing, to name but a few. De Beer et al. (1998) support the notion that tourism can lead to an unequal spread of benefits. Unemployment, which is a major problem in regions such as the Eastern Cape, Northern Province, Northern Cape, North West and the Free State, will thus not be alleviated with current tourism strategit"'i.
New strategies should consider that provinces need to develop their transport systems to make them more accessible particularly road networks and airports. This view is shared by Mkhize (1994). The Spatial Development Initiatives (SDIs), which should increase traffic to these provinces, can help in this regard (De Beer e/ al., 1998; Saayman & Saayman, 1999). Where possible, these provinces will also have to supply their own goods and fresh produce instead of "importing" them from Gauteng, Western Cape and KwaZulu-Natal in order to minimise leakages to these more affluent provinces. Provinces should also improve tourism-related services, for example recreation, entertainment, tour operators and tour guides, which will also benefit international tourism. Another element of a tourism strategy could be for provinces to focus primarily on the domestic market.
Looking at the effect of tourism on employment, the results in section 4 showed that the sectors benefiting most from tourism are manufacturing, transport and communications, catering and accommodation, services and agriculture. Of these, only catering and accommodation, services (non-financial) and agriculture can be classified as relatively labour-intensive industries (see Appendix B). Food and food processing and beverages (the manufacturing industries which benefit most) can be seen as moderately labour-intensive and relatively capital-intensive industries respectively (see Appendix B).

CONCLUSIONS
It has been argued in this paper that the spatial implications of tourism development might depend, inter alia, on the relative impacts of domestic versus international tourist spending. Using an input-output model, support was found for the notion that there should be greater emphasis on the development of the domestic tourism market in South Africa, as the macroeconomic effect is greater and more beneficial to sectors prominent in poorer provinces.
It was also found that regions such as Gauteng, Western Cape and KwaZulu-Natal benefit most from an increase in the number of international tourists to and in South Africa. for these provinces are suppliers of services and manufactured products. Thpy are also the gateways to South Africa. This situation leads to an unequal distribution of tourism benefits, where the provinces which need development and revenue the most, receive the least. The implication is that substantial path dependency effects will operate on a regional level in the case of international tourism. Improvements in infrastructure and transport services are vital if these effects are to improve.
It may therefore be concluded that regional governments (particularly of the poorer regions) in South Africa should focus more on domestic tourism as a shortterm tourism development strategy. Here the path-dependency effects may be smaller. However, they should complement this policy stance by increasing investment in transport services and infrastructure, as part of a medium-to longerterm strategy for the development of the international tourism market. APPENDIX A: