三篇關於經濟成長與觀光之非線性因果關係論文：縱橫平滑門檻方法的應用

Three essays on the nonlinear causality between economic growth and tourism: Applications of panel smooth threshold approaches

蕭瑞銘(Juei-Ming Hsiao)

指導教授 ： 吳博欽

中原大學/商學院/商學博士學位學程/博士(2021年)

https://doi.org/10.6840/cycu202100125

摘要

本博士論文包含三篇文章，試圖構建PSTR家族模型，包括：縱橫平滑轉換向量誤差修正模型、空間縱橫平滑轉換迴歸模型及縱橫平滑轉換向量自我迴歸模型，以檢視新發展的計量經濟方法與經濟環境對入境觀光與經濟成長之間因果關係的影響。
第一篇文章擴展縱橫平滑轉換向量誤差修正模型，以研究經濟成長與觀光的因果關係。該模型同時解決了非線性、異質性與內生性的估計問題。實證結果支持因果關係在長期與短期都是雙向的、非線性的、隨時間與國家而變動的。其次，實質利率在經濟成長與觀光之間的聯結具有門檻效果。高水準的實質利率導致經濟成長與觀光回復到其長期均衡的時間將更長，而在短期則會加強一個變數對另一個變數的正向貢獻。換言之，總體經濟環境與政策是影響門檻效果的關鍵因素。
第二篇文章構建一個空間縱橫平滑轉換迴歸模型，以估計台灣在1990 - 2016年期間的入境觀光及其相關的空間效應。該模型考慮觀光的非線性與空間依賴性的特性，並可以衡量利率對空間相依性及迴歸因子邊際貢獻的門檻效果。實證結果顯示，台灣入境觀光客呈現非線性的過程及正的空間相依性，且相依性是十六個來源國與台灣之間實質利差的遞增函數。明顯地，貨幣政策在影響台灣入境觀光客上扮演的重要的角色。此外，所得、相對價格水準、觀光基礎設施與修正後的地理距離也對入境觀光人數具有非線性的影響。
第三篇文章建立一個以金融科技指數（FTI）為轉換變數的縱橫平滑轉換向量自我迴歸模型，以探討22個OECD國家的入境觀光與經濟成長之間的因果關係。該模型解決非線性、異質性與內生性等估計問題。實證結果支持因果關係是非線性的、雙向的且隨時間與國家而變化的，決定於每個國家在每個時期的FTI。在大多數FTI下，當期入境觀光的成長將排擠下一期的經濟成長。當FTI高於門檻（81.27）時，當期的經濟成長將損害下一期的入境觀光成長。至於FTI低於門檻（81.27）的OECD國家，觀光公司應積極地擴大觀光投資，以分享前一期經濟成長所帶來的觀光紅利。對於FTI高於門檻的OECD國家，政府應採取適當的政策以減少經濟成長對觀光業的影響，而觀光業者則應在經濟衰退時期增加觀光投資。

關鍵字

縱橫平滑轉換向量誤差修正模型 ； 空間縱橫平滑轉換迴歸模型 ； 縱橫平滑轉換向量自我迴歸模型 ； 依時與國家變動 ； 內生性 ； 入境觀光 ； 經濟成長

並列摘要

This doctoral dissertation comprises three essays attempting to investigate the impact of new development in the econometric methods and economic environments on the causality between inbound tourism and economic growth by constructing PSTR family models, including panel smooth transition vector error correction model, spatial panel smooth transition regression model and panel smooth transition vector autoregressive model.
The first essay develops a panel smooth transition vector error correction model to investigate the economic growth-tourism causality. This model simultaneously resolves the estimation problems of nonlinearity, heterogeneity and endogeneity. Empirical results support that the causality is bi-directional, nonlinear, time- and country-varying in both the long run and short run. The real interest rate causes threshold effects on the link between growth and tourism. High levels of real interest rates lead to a longer time for the growth and tourism to return back to their long run equilibrium values; however, they strengthen the positive contribution from one of the variables to the other variable in the short run. Macroeconomic environment and policy are key factors that influence the threshold effects.
The second essay constructs a spatial panel smooth transition regression model to estimate Taiwan’s inbound tourism and its associated spatial effect during 1990-2016. The model considers the characteristics of the nonlinearity and spatial dependence in tourism and can measure the threshold effect of the interest rate on the spatial dependence and the marginal contributions of the regressors. Empirical results show that Taiwan’s inbound tourists display a nonlinear process and positive spatial dependence. The dependence is an increasing function of the real interest rate differential among sixteen origin countries and Taiwan. Evidently, monetary policy plays a crucial role in influencing Taiwan’s inbound arrivals. Moreover, income, relative price level, tourism infrastructure, and the revised geographical distance also have nonlinear impacts on arrivals.
The third essay establishes a panel smooth transition vector autoregressive model with a financial technology index (FTI) as a transition variable to explore the causality between inbound tourism and economic growth in 22 OECD countries. The model resolves the estimation problems of nonlinearity, heterogeneity, and endogeneity. The empirical results support that the causality is nonlinear, bidirectional, and time- and country-varying, depending on each country’s FTI in each period. Under most FTIs, the current inbound tourism growth will crowd out the economic growth in the next period. When the FTI is higher than the threshold (81.27), the current economic growth will harm inbound tourism growth in the next period. For OECD countries with an FTI below the threshold (81.27), tourism companies should actively expand their investment to share the tourism dividend driven from the economic growth in the previous period. For OECD countries with an FTI above the threshold, the governments should adopt appropriate policies to reduce the impact of economic growth on the tourism industry, and tourism companies should increase tourism investment during the periods of economic downturn.

並列關鍵字

Panel smooth transition vector error correction model ； spatial panel smooth transition regression model ； panel smooth transition vector autoregressive model ； time- and country-varying ； endogeneity ； inbound tourism ； economic growth

參考文獻

Anderson, D. R., Sweeney, D. J., Williams, T. A., Camm, J. D., Cochran, J. J. (2016). Statistics for Business Economics. Cengage Learning.

Ando, A., Modigliani, F. (1963). The life cycle hypothesis of saving. American Economic Review, 53 (1), 55-84.

Anselin, L. (1988). A test for spatial autocorrelation in seemingly unrelated regressions. Economics Letters, 28(4), 335-341.

Anselin, L. (1995). Local indicators of spatial association – LISA. Geographical Analysis, 27(2), 93-115.

Anselin, L. (2001). Spatial Econometrics. In B. H. Baltagi (Ed.), A Companion to Theoretical Econometrics (pp. 310-330). Massachusetts: Blackwell.