آیا توسعه مبتنی بر صنعت گردشگری برای فقرا خوب است؟ تجزیه و تحلیل تعادل عمومی برای تایلند

In low-income countries, the use of tax revenues to fund tourism promotions is motivated in part by the belief that tourism growth will improve income distribution by expanding demand for relatively low-skilled labor. We examine this belief for the case of Thailand, a highly tourism-intensive economy, using a new and specifically designed applied general equilibrium model. Thailand's tourism boom, fueled in part by a series of publicly funded promotional campaigns, has coincided with a period of worsening inequality. We find that growth of inbound tourism demand raises aggregate household income, but worsens its distribution. This is because tourism sectors are not especially labor-intensive in the Thai context, and because the expansion of foreign tourism demand creates general equilibrium effects that undermine profitability in tradable sectors (such as agriculture) from which the poor derive a substantial fraction of their income. These results indicate that tourism growth is not a panacea for other goals of development policy; to address inequality, additional policy instruments are required. We explore this implication with the example of a lump-sum tax imposed at different rates for rich and poor households. In addition, we examine the robustness of our main results with respect to alternative factor market assumptions relevant to the Thai economy.

Tourism is an increasingly popular component of development strategy in low-income countries. This popularity seems to be based on three beliefs. First, that tourism can serve as a substantial source of foreign exchange earnings, so contributing to economic growth. Second, that tourism services are labor-intensive, so expansion of this industry will create jobs and improve income distribution. Third, that tourism is a “clean” industry, i.e., its growth is good for the environment. The promotion of tourism thus appears to generate private gains and also to advance broader societal goals; in particular, policies that promote tourism are seen as ‘pro-poor’ in that they are supposed to create disproportionately more jobs for less-skilled (and thus poorer) workers. The first of these beliefs is well founded in many cases. In many developing countries travel and tourism (T&T) sectors contribute a larger share to total GDP than the world average and also generate a larger than average share of jobs and exports (WTTC, 2006).1 The other two beliefs are less robust, however, at least in the case of Thailand, a major tourist destination and a country in which tourism is large in relation to national aggregates. Tourism expansion in Thailand certainly creates jobs for unskilled workers, and this has a direct poverty alleviation impact. But much of the gain from tourism growth accrues to factors other than unskilled labor, so income distribution may actually worsen. In addition, low-skill jobs in other sectors may be destroyed, and returns to agricultural land, from which the poor derive a considerable share of their income, may fall as tourism expands.2 Government efforts to promote tourism growth may thus be inconsistent with the goal of reduced income inequality. When tourism is relatively large in relation to GDP and employment, changes affecting the industry have economy-wide impacts. Poverty and distributional outcomes of tourism growth cannot easily be predicted except in a numerical model capturing some of the complexity of interdependent sectors and markets, as well as the effects of economic policies and other distortions. Subject to availability of data, applied general equilibrium (AGE) models can be used to examine such complex economic systems. In this paper we present an AGE model for Thailand and simulate the effects of tourism growth. Our goal is to take account of general equilibrium adjustment mechanisms in answering the question: is tourism growth pro-poor?

The Thai economy depends heavily on the performance of its tourism industries. These account for millions of jobs and a substantial fraction of export earnings, and a wide range of other industries are directly or indirectly interdependent with them. Given this, any internal or external changes that affect Thai tourism can have economy-wide impacts on resource allocation, sectoral outputs, income distribution, and the macroeconomy. In this paper we have focused on the effects of accelerated inbound tourism growth on the distribution of factor and household incomes. Simulations with an AGE model reveal that although tourism growth benefits all household classes, the biggest gains accrue to high-income and non-agricultural households. In a relative sense, inbound tourism expansion is not a pro-poor change so long as the owners of primary factors in agriculture and other labor-intensive tradables sectors do not participate in tourism-related activities. Growth of inbound tourism induces the reallocation of primary factors toward domestic-oriented production and away from tradables sectors, notably agriculture. As real wages and capital returns are greater in non-agriculture than in agriculture, the structural changes induced by tourism growth tend to further widen intersectoral differences in wages and capital returns. Owners of land, the income of which is tied directly to the fortunes of agriculture, also lose. These distributional results arise because contrary to widespread assumptions, tourism in Thailand – and in low-income countries generally – is not a notably labor-intensive sector relative to key tradables sectors such as agriculture and labor-intensive manufacturing. The broad effects of tourism growth are the same no matter how elastic is factor supply. Its economic impacts are stronger as factor market constraints are relaxed. However, high-income non-farm households are the greatest winners under all factor market assumptions. We also find that the benefits of a tourism boom are spread widely across sectors, in contrast with typical Dutch Disease theory in which a boom is concentrated in one sector. The resource movement effect of tourism dominates its spending effect. This finding contrasts with Dutch Disease models of booms in oil, mining and other natural resource sectors, in which the effects of factor market linkages are relatively minor and the spending effect is dominant. Our analysis enables the generation of policy-relevant information on the role of tourism in a developing country, by breaking the sector out in a national accounting data set and incorporating those data in a model built specifically to accommodate structural features of a tourism-led economy. Importantly, our approach captures the general equilibrium constraints imposed by factor endowments, and these, it appears, play an important role in shaping changes in economic structure following an exogenous increase in tourism demand. This finding, plus the observation that tourism growth per se does not raise unskilled wages or the incomes of the poor relative to the incomes of the rich, generates some practical policy insights. First, as with any distortionary measure, a tourism promotion campaign – that is, a subsidy to tourism industries – is an implicit redistributive policy, and in the Thai case worsens the gap between rich and poor. Second, it follows that if the government has more than one policy goal, it must deploy more than one instrument. Promoting tourism is not, by itself, a panacea for other goals of development policy. Programs of transfers to the rural poor, most recently those initiated by the government of Thaksin Shinawatra (2001–2006), may provide partial compensation, but the trend in the national Gini ratio suggests that more is needed, given current Thai growth policy settings, if inequality is really to be reduced. Finally, in resolving one puzzle, the research on which this paper is based has generated new empirical and policy questions and poses new methodological challenges. As with most models of its kind, the sectoral and structural richness of our model is a characteristic won at the expense of other features, notably dynamics and second-moment effects. Might tourism growth reduce the pace of long-term human capital accumulation by drawing skilled workers and entrepreneurs away from ‘cutting-edge’ manufacturing industries such as electronics? Does the volatility of the industry, and in particular its vulnerability to exogenous shocks due to weather, disease outbreaks, fears of terrorist attack, and perceptions of political and economic stability, give rise to transactions costs that reduce its measured contribution to welfare growth? What are the net environmental and poverty impacts of tourism, in an economy in which natural resource wealth remains the primary income source for the majority of the poor? These are questions for which the increasing availability of richer data sets should in the near future justify the development and application of more complex models.