(Luận văn) the mutual effects of shadow economy and financial development in asean countries

INSTITUTE OF SOCIAL STUDIES UNIVERSITY OF ECONOMICS ERASMUS UNIVERSITY ROTTERDAM HO CHI MINH CITY THE HAGUE VIETNAM THE NETHERLAND t to SCHOOL OF ECONOMICS ng hi ep w n lo ad IN DEVELOPMENT ECONOMICS ju y th VIETNAM - THE NETHERLANDS PROGRAMME FOR M.A yi pl n ua al n va THE MUTUAL EFFECTS OF SHADOW ECONOMY AND ll fu FINANCIAL DEVELOPMENT IN ASEAN COUNTRIES oi m at nh z by Nguyen Hoang Phu z k jm the degree of ht vb A thesis submitted in partial fulfilment of the requirements for om Development Economics l.c gm Master of Art in an Lu n va ey t re Ho Chi Minh city, January 2018 t to ng hi VIETNAM - THE NETHERLANDS PROGRAMME FOR M.A IN ep DEVELOPMENT ECONOMICS w n lo ad THE MUTUAL EFFECTS OF SHADOW ECONOMY AND y th ju FINANCIAL DEVELOPMENT IN ASEAN COUNTRIES yi pl n ua al va n by Nguyen Hoang Phu ll fu oi m A thesis submitted in partial fulfilment of the requirements for at nh the degree of z z Development Economics k jm ht vb Master of Art in om an Lu Dr Pham Thi Thu Tra l.c gm Academic Supervisor: n va ey t re Ho Chi Minh city, January 2018 DECLARATION t to I hereby declare that my dissertation entitled “THE MUTUAL EFFECTS OF ng SHADOW ECONOMY AND FINANCIAL DEVELOPMENT IN ASEAN hi COUNTRIES” is the result of my own work and includes nothing which is the outcome ep of work done in collaboration except as declared in the Preface and specified in the text w I also confirm that: n lo  This thesis was done wholly while in candidature for a research degree at ad y th VNP; ju  Where any part of this thesis has previously been submitted for a degree or yi any other qualification at VNP or any other institution, this has been clearly pl ua al stated; n  Where I have consulted the published work of others, this is always clearly n va attributed; ll fu  Where I have quoted from the work of others, the source is always given oi m  With the exception of such quotations, this thesis is entirely my own work, at nh and I have acknowledged all main sources of help z z k jm ht vb l.c gm Date: January 02, 2018 Signature om Full name: Nguyen Hoang Phu an Lu n va ey t re ACKNOWLEDGEMENT This thesis cannot complete without the support of my supervisor, Dr Pham Thi t to Thu Tra, who has spent the value time, efforts, and energy to guide me on the thesis ng during the time of completing the thesis Her dedication made me motivated when I hi have a chance to discuss with her, her expertise is what makes me impressive when I ep ask her questions about my thesis’s topic, and she also kept me in the “can – do” attitude w when I faced any difficulties in doing thesis All of these leave me with the most n lo unforgettable memory and experience My purpose of this acknowledgement is to ad y th express my gratitude to my supervisor Without her supports, I may not have a chance ju to pursue my dream yi I would like to send my special thanks to Prof Nguyen Trong Hoai, Dr Pham pl ua al Khanh Nam, Dr Truong Dang Thuy for their valuable command, guidance and support n during the program Without your support and encouragement, I may not complete the n va thesis as expected ll fu Additionally, my thanks are given to all of the lectures who have been my oi m knowledge guiders and the staff who have been my service supporters throughout the nh master program at University of Economics and Erasmus University Rotterdam at Without their help, never can I have an opportunity to proceed and complete my master z z thesis vb jm ht Last but not least, I would like to thank Mr Nguyen Cong Thanh, Truong Thi Thu May and my family who have always been a pillar for me to rely on during the hardships k gm of attempting to achieve the master thesis It is their unspoken sacrifice and untiring om progress l.c work that bring me more spare time to be able to reach the final destination of my an Lu n va ey t re ABSTRACT This study focuses on examining the mutual relationship between financial t to development and shadow economy by applying the theoretical and empirical ng framework Our research contributed to the way of calculating the size of shadow hi economy applied the currency demand approach with updated data from 1997 to 2015 ep for ASIAN countries In particular, to have a robust result, we used estimation w methods including POLS, FEM, REM and SGMM to calculate the value of the size of n lo shadow economy of each country Then, we took each received results to examine the ad y th mutual effect with the financial development using P – VAR approach We found that ju when the positive shock caused by the financial sector affects the shadow economy, the yi shadow economy will immediately respond negatively to the shock On the other hand, pl ua al when a positive shock caused by credit for private sector will lead to the positive n responses of the shadow economy Interestingly, in this case, the response tends to last va longer with the estimated results from static model of shadow economy in comparison n ll fu with dynamic model of shadow economy oi m at JEL classifications: G32, H26 nh Keywords: Shadow economy, financial development z z k jm ht vb om l.c gm an Lu n va ey t re TABLE OF CONTENTS t to Chapter 1: Introduction 11 ng hi ep 1.1 Problem statements 11 1.2 Research objectives 13 1.3 Scope of the study 14 w n lo 1.4 Structure of the thesis 15 ad Chapter 2: Literature review 16 y th 2.1 Review of theory 16 ju The theory of shadow economy 16 2.1.2 The review on financial development theories 21 pl ua al 2.2 yi 2.1.1 Review of empirical studies on the relationship between financial development n Summary 31 ll fu 2.3 n va theory and shadow economy theory 27 oi m Chapter 3: Research methodology 34 Analytical framework 34 3.2 Econometric models 36 3.3 Data 40 3.4 Variables and sampling 40 at nh 3.1 z z jm ht vb k Chapter 4: Research results 42 gm Overview of the research topic 42 4.2 Descriptive statistics 44 4.3 Regression results and discussions 45 om l.c 4.1 an Lu Chapter 5: Conclusions 55 Limits of the study 56 Appendices 65 ey Reference 58 t re 5.2 n Conclusions 55 va 5.1 t to LIST OF TABLES ng hi ep Table 1: Descriptive Statistics for the whole dataset 44 w Table 2: Matrix of correlation coefficients 44 n lo Table 3: Estimated results of currency demand model 46 ad Table 4: Optimal model selection tests 47 y th Table 5: Estimated value of Shadow economy over GDP 48 ju yi Table 6: The results of Unit Root Test 49 pl n ua al n va ll fu oi m at nh z z k jm ht vb om l.c gm an Lu n va ey t re LIST OF CHARTS Figure 1: Conceptual framework of shadow economy and financial development 31 Figure 2: The technical structure to deal with data 34 t to Figure 3: The analyzed results of impulse response function when the size of shadow ng hi economic creates a shock with the estimated value of shadow economy from the ep static models (POLS, FEM, REM) 51 Figure 4: The analyzed results of impulse response function when the size of shadow w n economic creates a shock with the estimated value of shadow economy from the lo ad dynamic models (SGMM) 52 y th Figure 5: The analyzed results of impulse response function when the financial development ju creates a shock with the estimated value of shadow economy from the static models yi pl (PLOS, FEM, REM) 53 ua al Figure 6: The analyzed results of impulse response function when the financial development n creates a shock with the estimated value of shadow economy from the dynamic n va models (SGMM)) 54 ll fu oi m at nh z z k jm ht vb om l.c gm an Lu n va ey t re LIST OF APPENDICES Appendix 1: Shadow economy estimation using POLS 65 t to ng Appendix 2: Shadow economy estimation using FEM 65 hi Appendix 3: Shadow economy estimation using SGMM 66 ep Appendix 4: Shadow economy estimation using REM 67 w Appendix 5: Hausman Test 67 n Appendix 6: Breusch & Pagan Lagrangian multiplier test for random effects 68 lo ad Appendix 7: Stationary test for shadow size estimated by POLS 68 ju y th Appendix 8: Stationary test for first difference of shadow size estimated by POLS 69 Appendix 9: Stationary test for shadow size estimated by FEM 69 yi pl Appendix 10: Stationary test for first difference of shadow size estimated by FEM 70 ua al Appendix 11: Stationary test for shadow size estimated by REM 71 n Appendix 12: Stationary test for first difference of shadow size estimated by REM 71 va Appendix 13: Stationary test for shadow size estimated by SGMM 72 n ll fu Appendix 14: Stationary test for first difference of shadow size estimated by SGMM 72 oi m Appendix 15: Stationary test for financial development measured by Credit for private sectors 73 nh at Appendix 16: Stationary test for first difference of financial development measured by Credit z for private sectors 73 z ht vb Appendix 17: Stationary test for financial development measured by Credit from financial jm sectors 74 k Appendix 18: Stationary test for first difference of financial development measured by Credit gm from financial sectors 74 l.c Appendix 19: Stationary test for money supply ratio 75 om Appendix 20: Stationary test for first difference of money supply ratio 75 an Lu Appendix 21: Stationary test for natural logarithm of GDP per capita 76 Appendix 22: Stationary test for first difference of natural logarithm of GDP per capita 76 n va ey t re ABBREVIATIONS t to ng hi POLS: Pooled Ordinary Least Squared FEM: Fixed Effects Model REM: Random Effects Model ep w n lo ad ju y th yi pl n ua al n va ll fu oi m at nh z z k jm ht vb om l.c gm an Lu n va ey t re Isham, J., Woolcock, M., Pritchett, L., & Busby, G (2005) The varieties of resource experience: natural resource export structures and the political economy of economic growth The World Bank Economic Review, 19(2), 141-174 Jaffee, D., & Levonian, M (2001) The structure of banking systems in developed and transition t to economies European Financial Management, 7(2), 161-181 ng Johnson, S., Kaufmann, D., & Shleifer, A (1997) Politics and entrepreneurship in transition hi ep economies Johnson, S., Kaufmann, D., & Zoido-Lobaton, P (1998) Regulatory discretion and the unofficial w n economy The American Economic Review, 88(2), 387-392 lo ad Khan, A (2001) Financial development and economic growth Macroeconomic dynamics, 5(3), y th 413-433 ju Klovland, J T (1984) Tax evasion and the demand for currency in Norway and Sweden Is there yi a hidden relationship? The Scandinavian Journal of Economics, 423-439 pl al Lê Đăng Doanh, & Nguyễn Minh Tú (1997) Khu vực kinh tế phi quy - Một số kinh nghiệm n va Nội) n ua quốc tế thực tiễn VN q trình chuyển đổi kinh tế (NXB Chính trị quốc gia, Hà Levin, A., Lin, C.-F., & Chu, C.-S J (2002) Unit root tests in panel data: asymptotic and finite- fu ll sample properties Journal of econometrics, 108(1), 1-24 m oi Levine, R (1997) Financial development and economic growth: views and agenda Journal of nh economic literature, 35(2), 688-726 at z Love, I., & Zicchino, L (2006) Financial development and dynamic investment behavior: z vb Evidence from panel VAR The Quarterly Review of Economics and Finance, 46(2), 190- jm ht 210 Lütkepohl, H (2005) New introduction to multiple time series analysis: Springer Science & k gm Business Media l.c Maddala, G S., & Wu, S (1999) A comparative study of unit root tests with panel data and a om new simple test Oxford Bulletin of Economics and statistics, 61(S1), 631-652 an Lu Mai, H., & Schneider, F (2016) Modelling the Egyptian Shadow Economy: A MIMIC model and A Currency Demand approach Journal of Economics and Political Economy, 3(2), ey Economics, 90(2), 163-178 t re Malik, A., & Temple, J R (2009) The geography of output volatility Journal of Development n va 309 62 Martimort, D., & Straub, S (2005) The Political Economy of Private Participation, Social Discontent and Regulatory Governance Background paper commissioned for this Report Inter-American Development Bank, Washington, DC Mayer, C., & Sussman, O (2001) The assessment: finance, law, and growth Oxford Review of t to Economic Policy, 17(4), 457-466 ng McKinnon, R., & Shaw, E (1973) Financial deepening in economic development Washington, hi ep Brookings Institution Medina, L., & Schneider, F (2017) Shadow Economies around the World: New Results for 158 w n Countries over 1991-2015 lo ad Nguyễn Thị Thùy Linh, Trần Huy Hoàng, & Nguyễn Hữu Huân (2015) Mối quan hệ tự y th hố tài đến tăng trưởng kinh tế quốc gia nổi: Trường hợp Việt Nam Tạp ju chí Phát triển Kinh tế, 26(10), 02-26 yi Pagano, M., & Volpin, P (2001) The political economy of finance Oxford Review of Economic pl ua al Policy, 17(4), 502-519 Porta, R L., Lopez-de-Silanes, F., Shleifer, A., & Vishny, R W (1998) Law and finance n n va Journal of political economy, 106(6), 1113-1155 Qin, D (2011) Rise of VAR modelling approach Journal of Economic Surveys, 25(1), 156-174 fu ll Rajan, R G., & Zingales, L (2003) The great reversals: the politics of financial development in m oi the twentieth century Journal of financial economics, 69(1), 5-50 nh Sachs, J D., Warner, A., Åslund, A., & Fischer, S (1995) Economic reform and the process of at z global integration Brookings papers on economic activity, 1995(1), 1-118 z jm ht European Economic Review, 36(4), 763-781 vb Saint-Paul, G (1992) Technological choice, financial markets and economic development Schneider, F (1986) Estimating the size of the Danish shadow economy using the currency k gm demand approach: An attempt The Scandinavian Journal of Economics, 643-668 l.c Schneider, F (1994) Measuring the size and development of the shadow economy Can the om causes be found and the obstacles be overcome? Essays on economic psychology (pp an Lu 193-212): Springer Schneider, F (2002) Size and measurement of the informal economy in 110 countries Paper 63 ey Schneider, F (2011) Handbook on the shadow economy: Edward Elgar Publishing t re Journal of Political Economy, 21(3), 598-642 n Schneider, F (2005) Shadow economies around the world: what we really know? European va presented at the Workshop of Australian National Tax Centre, ANU, Canberra Schneider, F., Buehn, A., & Montenegro, C E (2010) New estimates for the shadow economies all over the world International Economic Journal, 24(4), 443-461 Schneider, F., & Enste, D H (2000) Shadow economies: size, causes, and consequences Journal of economic literature, 38(1), 77-114 t to Schneider, F., & Enste, D H (2013) The shadow economy: An international survey: Cambridge ng University Press hi ep Scott Hacker, R., & Hatemi-J, A (2008) Optimal lag-length choice in stable and unstable VAR models under situations of homoscedasticity and ARCH Journal of Applied Statistics, w n 35(6), 601-615 lo ad Shaw, E S (1973) Financial deepening in economic development y th Smith, P M (1997) Assessing the size of the underground economy: The statistics Canada ju perspective The underground economy: Global evidence of its size and impact, 11-37 yi Stiglitz, J E., & Stiglitz, J E (2000) Economics of the public sector pl 78(2), 299-321 n ua al Straub, S (2005) Informal sector: the credit market channel Journal of Development Economics, economics, 70(3), 313-349 n va Stulz, R M., & Williamson, R (2003) Culture, openness, and finance Journal of financial fu ll Tanzi, V (1982) The underground economy in the United States and abroad: Free Press m oi Tanzi, V (1983) The underground economy in the United States: Annual estimates, 1930-80 nh (L'économie clandestine aux Etats-Unis: estimations annuelles, 1930-80)(La" economía at z subterránea" de Estados Unidos: Estimaciones anuales, 1930-80) Staff Papers- z vb International Monetary Fund, 283-305 jm ht Teobaldelli, D (2011) Federalism and the shadow economy Public Choice, 146(3), 269-289 Torgler, B., & Schneider, F (2009) The impact of tax morale and institutional quality on the k gm shadow economy Journal of Economic Psychology, 30(2), 228-245 l.c Võ Hồng Đức, Lý Hưng Thịnh, & Tống Thị Hồng Nhung (2015) Kinh tế ngầm tỉ lệ thất om nghiệp: Bằng chứng thực nghiệm quốc gia Châu Á Tạp chí Cơng nghệ Ngân an Lu hàng(109), Vũ Việt Quảng, & Lê Thị Phương Vy (2016) Mối quan hệ thể chế khả tiếp cận thị n va trường vốn doanh nghiệp Tạp chí Phát triển Kinh tế, 27(6), 80- 101 ey t re Wooldridge, J M (2015) Introductory econometrics: A modern approach Nelson Education 64 Appendices Appendix 1: Shadow economy estimation using POLS t to Source SS df MS ng hi 9.70905237 10.3589491 88 2.42726309 117715331 Total 20.0680015 92 218130451 ep Model Residual Number of obs F(4, 88) Prob > F R-squared Adj R-squared Root MSE = = = = = = 93 20.62 0.0000 0.4838 0.4603 3431 w n ln_mic_m2 Coef Std Err t P>|t| [95% Conf Interval] lo ad ju y th ln_gov_tax_gni_p1 ln_pri_exp_gni ln_dep_int ln_gdp_per_capita _cons yi 2.921145 -.7560456 2572766 -.2064133 -.38018 4707391 166444 0573786 0423472 355015 6.21 -4.54 4.48 -4.87 -1.07 0.000 0.000 0.000 0.000 0.287 1.98565 -1.086818 1432487 -.2905694 -1.085698 3.85664 -.4252731 3713045 -.1222571 3253378 pl ua al Appendix 2: Shadow economy estimation using FEM Number of obs Number of groups n Fixed-effects (within) regression Group variable: cross n va ll fu oi m R-sq: within = 0.3813 between = 0.1580 overall = 0.2025 93 Obs per group: = avg = max = 13.3 16 F(4,82) Prob > F = = nh corr(u_i, Xb) = = = -0.3946 12.64 0.0000 at t ln_gov_tax_gni_p1 ln_pri_exp_gni ln_dep_int ln_gdp_per_capita _cons 2.064902 -.539919 -.1719121 -.3959178 1.663089 5217305 1703693 0333121 0970453 7574193 sigma_u sigma_e rho 46932269 13393875 92468794 (fraction of variance due to u_i) [95% Conf Interval] 0.000 0.002 0.000 0.000 0.031 1.027014 -.8788378 -.2381805 -.5889718 1563407 k jm ht 3.10279 -.2010002 -.1056437 -.2028638 3.169837 l.c gm F test that all u_i=0: F(6, 82) = 82.57 vb 3.96 -3.17 -5.16 -4.08 2.20 P>|t| om Coef z Std Err z ln_mic_m2 Prob > F = 0.0000 an Lu n va ey t re 65 Appendix 3: Shadow economy estimation using SGMM Dynamic panel-data estimation, two-step system GMM t to Group variable: cross Time variable : years Number of instruments = 31 Wald chi2(5) = 44.13 Prob > chi2 = 0.000 Number of obs Number of groups Obs per group: avg max ng z P>|z| 88 12.57 15 ln_mic_m2 Coef ln_mic_m2 L1 -.3916444 9166027 -0.43 0.669 -2.188153 1.404864 ln_gov_tax_gni_p1 ln_pri_exp_gni ln_dep_int ln_gdp_per_capita _cons 4.583396 -1.340635 -.4540552 -1.284231 8.304857 8.676479 1.699513 2524483 1.289465 10.27017 0.53 -0.79 -1.80 -1.00 0.81 0.597 0.430 0.072 0.319 0.419 -12.42219 -4.671618 -.9488448 -3.811536 -11.8243 21.58898 1.990349 0407345 1.243073 28.43401 hi Std Err = = = = = [95% Conf Interval] ep w n lo ad y th ju Warning: Uncorrected two-step standard errors are unreliable yi pl Instruments for first differences equation Standard D.(ln_gov_tax_gni_p1 ln_pri_exp_gni ln_dep_int ln_gdp_per_capita) GMM-type (missing=0, separate instruments for each period unless collapsed) L2.L.ln_mic_m2 Instruments for levels equation Standard ln_gov_tax_gni_p1 ln_pri_exp_gni ln_dep_int ln_gdp_per_capita _cons GMM-type (missing=0, separate instruments for each period unless collapsed) DL.L.ln_mic_m2 n ua al n va ll fu oi m at nh Arellano-Bond test for AR(1) in first differences: z = Arellano-Bond test for AR(2) in first differences: z = z overid restrictions: chi2(25) = 24.68 but not weakened by many instruments.) overid restrictions: chi2(25) = 0.13 weakened by many instruments.) Pr > z = Pr > z = 0.716 0.886 Prob > chi2 = 0.480 z vb Prob > chi2 = 1.000 jm ht Sargan test of (Not robust, Hansen test of (Robust, but -0.36 0.14 k Difference-in-Hansen tests of exogeneity of instrument subsets: GMM instruments for levels Hansen test excluding group: chi2(12) = 0.13 Prob > chi2 Difference (null H = exogenous): chi2(13) = -0.00 Prob > chi2 iv(ln_gov_tax_gni_p1 ln_pri_exp_gni ln_dep_int ln_gdp_per_capita) Hansen test excluding group: chi2(21) = 0.02 Prob > chi2 Difference (null H = exogenous): chi2(4) = 0.11 Prob > chi2 gm 1.000 1.000 = = 1.000 0.999 om l.c = = an Lu n va ey t re 66 Appendix 4: Shadow economy estimation using REM t to ng hi ep Random-effects GLS regression Group variable: cross Number of obs Number of groups = = 93 R-sq: within = 0.3764 between = 0.1902 overall = 0.2160 Obs per group: = avg = max = 13.3 16 corr(u_i, X) Wald chi2(4) Prob > chi2 = (assumed) = = 42.58 0.0000 w n ln_mic_m2 Coef Std Err z P>|z| [95% Conf Interval] lo ad ju y th ln_gov_tax_gni_p1 ln_pri_exp_gni ln_dep_int ln_gdp_per_capita _cons yi 2.039711 -.4038426 -.1472631 -.2963167 8737868 4.44 -2.80 -4.22 -4.01 1.49 0.000 0.005 0.000 0.000 0.135 1.139587 -.6861398 -.2156023 -.4412751 -.272933 2.939835 -.1215455 -.0789239 -.1513583 2.020507 pl 21887821 13393875 72755756 (fraction of variance due to u_i) n ua al sigma_u sigma_e rho 4592554 1440318 0348676 0739597 5850719 n va Appendix 5: Hausman Test sqrt(diag(V_b-V_B)) S.E .0251917 -.1360763 -.024649 -.0996011 2475626 0909974 0628311 at nh z 2.039711 -.4038426 -.1472631 -.2963167 (b-B) Difference oi 2.064902 -.539919 -.1719121 -.3959178 m ln_gov_tax~1 ln_pri_exp~i ln_dep_int ln_gdp_per~a ll fu Coefficients (b) (B) fem rem z Test: Ho: difference in coefficients not systematic k om l.c gm chi2(4) = (b-B)'[(V_b-V_B)^(-1)](b-B) = 0.80 Prob>chi2 = 0.9382 (V_b-V_B is not positive definite) jm ht vb b = consistent under Ho and Ha; obtained from xtreg B = inconsistent under Ha, efficient under Ho; obtained from xtreg an Lu n va ey t re 67 Appendix 6: Breusch & Pagan Lagrangian multiplier test for random effects Breusch and Pagan Lagrangian multiplier test for random effects ln_mic_m2[cross,t] = Xb + u[cross] + e[cross,t] Estimated results: Var t to ng ln_mic_m2 e u hi ep Test: sd = sqrt(Var) 2181305 0179396 0479077 4670444 1339387 2188782 Var(u) = chibar2(01) = Prob > chibar2 = 157.82 0.0000 w n Appendix 7: Stationary test for shadow size estimated by POLS lo ad xtunitroot fisher shadow_pols_gdp, dfuller lag(1) trend (40 missing values generated) y th ju Fisher-type unit-root test for shadow_pols_gdp Based on augmented Dickey-Fuller tests yi = Number of panels Avg number of periods = pl Ho: All panels contain unit roots Ha: At least one panel is stationary ua al Panel-specific Included Included Not included Asymptotics: T -> Infinity n va ADF regressions: lag n AR parameter: Panel means: Time trend: Drift term: fu p-value ll Statistic 29.1074 -2.2865 -2.4578 2.8550 oi 0.0101 0.0111 0.0093 0.0022 at nh P Z L* Pm m Inverse chi-squared(14) Inverse normal Inverse logit t(39) Modified inv chi-squared 13.29 z z P statistic requires number of panels to be finite Other statistics are suitable for finite or infinite number of panels k jm ht vb om l.c gm an Lu n va ey t re 68 Appendix 8: Stationary test for first difference of shadow size estimated by POLS xtunitroot fisher d.shadow_pols_gdp, dfuller lag(1) trend (47 missing values generated) Fisher-type unit-root test for D.shadow_pols_gdp Based on augmented Dickey-Fuller tests t to ng hi ep Number of panels = Avg number of periods = AR parameter: Panel means: Time trend: Drift term: Asymptotics: T -> Infinity Ho: All panels contain unit roots Ha: At least one panel is stationary w Panel-specific Included Included Not included ADF regressions: lag n lo ad ju y th Inverse chi-squared(14) Inverse normal Inverse logit t(39) Modified inv chi-squared 12.29 Statistic p-value 63.1031 -5.0914 -6.3735 9.2796 0.0000 0.0000 0.0000 0.0000 P Z L* Pm yi pl P statistic requires number of panels to be finite Other statistics are suitable for finite or infinite number of panels n ua al n va ll fu oi m nh Appendix 9: Stationary test for shadow size estimated by FEM at z xtunitroot fisher shadow_fe_gdp, dfuller lag(1) trend (40 missing values generated) z jm ht vb Fisher-type unit-root test for shadow_fe_gdp Based on augmented Dickey-Fuller tests Number of panels = Avg number of periods = AR parameter: Panel means: Time trend: Drift term: Asymptotics: T -> Infinity k Ho: All panels contain unit roots Ha: At least one panel is stationary 69 ey P statistic requires number of panels to be finite Other statistics are suitable for finite or infinite number of panels t re 0.4987 0.2527 0.2756 0.5484 n 13.3563 -0.6659 -0.6011 -0.1217 va p-value an Lu Statistic om P Z L* Pm ADF regressions: lag l.c gm Panel-specific Included Included Not included Inverse chi-squared(14) Inverse normal Inverse logit t(39) Modified inv chi-squared 13.29 Appendix 10: Stationary test for first difference of shadow size estimated by FEM xtunitroot fisher d.shadow_fe_gdp, dfuller lag(1) trend (47 missing values generated) t to Fisher-type unit-root test for D.shadow_fe_gdp Based on augmented Dickey-Fuller tests ng hi ep Number of panels = Avg number of periods = AR parameter: Panel means: Time trend: Drift term: Asymptotics: T -> Infinity Ho: All panels contain unit roots Ha: At least one panel is stationary w n Panel-specific Included Included Not included 12.29 ADF regressions: lag lo ad y th ju Inverse chi-squared(14) Inverse normal Inverse logit t(39) Modified inv chi-squared yi Statistic p-value 86.2281 -5.2346 -8.5720 13.6498 0.0000 0.0000 0.0000 0.0000 P Z L* Pm pl n ua al P statistic requires number of panels to be finite Other statistics are suitable for finite or infinite number of panels n va ll fu oi m at nh z z k jm ht vb om l.c gm an Lu n va ey t re 70 Appendix 11: Stationary test for shadow size estimated by REM xtunitroot fisher shadow_re_gdp, dfuller lag(1) trend (40 missing values generated) Fisher-type unit-root test for shadow_re_gdp Based on augmented Dickey-Fuller tests t to ng hi ep Number of panels = Avg number of periods = AR parameter: Panel means: Time trend: Drift term: Asymptotics: T -> Infinity Ho: All panels contain unit roots Ha: At least one panel is stationary w Panel-specific Included Included Not included ADF regressions: lag n lo ad ju y th Inverse chi-squared(14) Inverse normal Inverse logit t(39) Modified inv chi-squared 13.29 Statistic p-value 13.6674 -0.6890 -0.6247 -0.0628 0.4748 0.2454 0.2679 0.5251 P Z L* Pm yi pl P statistic requires number of panels to be finite Other statistics are suitable for finite or infinite number of panels n ua al va Appendix 12: Stationary test for first difference of shadow size estimated by REM n ll fu xtunitroot fisher d.shadow_re_gdp, dfuller lag(1) trend (47 missing values generated) m oi Fisher-type unit-root test for D.shadow_re_gdp Based on augmented Dickey-Fuller tests nh Number of panels = Avg number of periods = AR parameter: Panel means: Time trend: Drift term: Asymptotics: T -> Infinity at Ho: All panels contain unit roots Ha: At least one panel is stationary z z jm ht ADF regressions: lag 0.0000 0.0000 0.0000 0.0000 an Lu P statistic requires number of panels to be finite Other statistics are suitable for finite or infinite number of panels om 90.6858 -5.5869 -9.1158 14.4922 l.c p-value gm Statistic k P Z L* Pm vb Panel-specific Included Included Not included Inverse chi-squared(14) Inverse normal Inverse logit t(39) Modified inv chi-squared 12.29 n va ey t re 71 Appendix 13: Stationary test for shadow size estimated by SGMM xtunitroot fisher shadow_gmm_gdp, dfuller lag(1) trend (40 missing values generated) t to Fisher-type unit-root test for shadow_gmm_gdp Based on augmented Dickey-Fuller tests ng hi ep Number of panels = Avg number of periods = AR parameter: Panel means: Time trend: Drift term: Asymptotics: T -> Infinity Ho: All panels contain unit roots Ha: At least one panel is stationary w n Panel-specific Included Included Not included 13.29 ADF regressions: lag lo ad y th ju Inverse chi-squared(14) Inverse normal Inverse logit t(39) Modified inv chi-squared yi Statistic p-value 79.4185 -3.5499 -7.6736 12.3629 0.0000 0.0002 0.0000 0.0000 P Z L* Pm pl n ua al P statistic requires number of panels to be finite Other statistics are suitable for finite or infinite number of panels n va ll fu Appendix 14: Stationary test for first difference of shadow size estimated by SGMM oi m Fisher-type unit-root test for D.shadow_gmm_gdp Based on augmented Dickey-Fuller tests nh Number of panels = Avg number of periods = AR parameter: Panel means: Time trend: Drift term: Asymptotics: T -> Infinity at Ho: All panels contain unit roots Ha: At least one panel is stationary z vb ADF regressions: lag an Lu P statistic requires number of panels to be finite Other statistics are suitable for finite or infinite number of panels om 0.0000 0.0009 0.0000 0.0000 l.c 55.3747 -3.1128 -5.1191 7.8191 gm p-value k Statistic jm ht P Z L* Pm z Panel-specific Included Included Not included Inverse chi-squared(14) Inverse normal Inverse logit t(39) Modified inv chi-squared 12.29 n va ey t re 72 Appendix 15: Stationary test for financial development measured by Credit for private sectors xtunitroot fisher credit_private, dfuller lag(1) trend (2 missing values generated) t to Fisher-type unit-root test for credit_private Based on augmented Dickey-Fuller tests ng hi ep Number of panels = Avg number of periods = AR parameter: Panel means: Time trend: Drift term: Asymptotics: T -> Infinity Ho: All panels contain unit roots Ha: At least one panel is stationary w n Panel-specific Included Included Not included 18.71 ADF regressions: lag lo ad y th ju Inverse chi-squared(14) Inverse normal Inverse logit t(39) Modified inv chi-squared yi Statistic p-value 88.4868 -1.3597 -5.8852 14.0767 0.0000 0.0870 0.0000 0.0000 P Z L* Pm pl n ua al P statistic requires number of panels to be finite Other statistics are suitable for finite or infinite number of panels n va ll fu m Appendix 16: Stationary test for first difference of financial development measured by Credit for private oi sectors at nh xtunitroot fisher d.credit_private, dfuller lag(1) trend (9 missing values generated) z z Fisher-type unit-root test for D.credit_private Based on augmented Dickey-Fuller tests vb Number of panels = Avg number of periods = AR parameter: Panel means: Time trend: Drift term: Asymptotics: T -> Infinity ADF regressions: lag 0.0000 0.0000 0.0000 0.0000 n ey t re P statistic requires number of panels to be finite Other statistics are suitable for finite or infinite number of panels va 113.9170 -6.4883 -11.7024 18.8825 an Lu p-value om Statistic l.c P Z L* Pm gm Panel-specific Included Included Not included Inverse chi-squared(14) Inverse normal Inverse logit t(39) Modified inv chi-squared 17.71 k jm ht Ho: All panels contain unit roots Ha: At least one panel is stationary 73 Appendix 17: Stationary test for financial development measured by Credit from financial sectors xtunitroot fisher credit_financial, dfuller lag(1) trend (2 missing values generated) Fisher-type unit-root test for credit_financial Based on augmented Dickey-Fuller tests t to ng hi ep Number of panels = Avg number of periods = AR parameter: Panel means: Time trend: Drift term: Asymptotics: T -> Infinity Ho: All panels contain unit roots Ha: At least one panel is stationary w Panel-specific Included Included Not included ADF regressions: lag n lo ad ju y th Inverse chi-squared(14) Inverse normal Inverse logit t(34) Modified inv chi-squared 18.71 Statistic p-value 6.4830 2.3695 2.5651 -1.4206 0.9528 0.9911 0.9926 0.9223 P Z L* Pm yi pl P statistic requires number of panels to be finite Other statistics are suitable for finite or infinite number of panels n ua al n va fu Appendix 18: Stationary test for first difference of financial development measured by Credit from financial ll sectors m oi xtunitroot fisher d.credit_financial, dfuller lag(1) trend (9 missing values generated) nh at Fisher-type unit-root test for D.credit_financial Based on augmented Dickey-Fuller tests z z Number of panels = Avg number of periods = AR parameter: Panel means: Time trend: Drift term: Asymptotics: T -> Infinity 17.71 jm ht vb Ho: All panels contain unit roots Ha: At least one panel is stationary ADF regressions: lag 47.2937 -4.3308 -4.8123 6.2919 0.0000 0.0000 0.0000 0.0000 n va P statistic requires number of panels to be finite Other statistics are suitable for finite or infinite number of panels an Lu p-value om Statistic l.c P Z L* Pm gm Inverse chi-squared(14) Inverse normal Inverse logit t(39) Modified inv chi-squared k Panel-specific Included Included Not included ey t re 74 Appendix 19: Stationary test for money supply ratio xtunitroot fisher m2_gdp , dfuller lag(1) trend (40 missing values generated) t to Fisher-type unit-root test for m2_gdp Based on augmented Dickey-Fuller tests ng hi ep Number of panels = Avg number of periods = AR parameter: Panel means: Time trend: Drift term: Asymptotics: T -> Infinity Ho: All panels contain unit roots Ha: At least one panel is stationary w n Panel-specific Included Included Not included 18.71 ADF regressions: lag lo ad y th ju Inverse chi-squared(14) Inverse normal Inverse logit t(39) Modified inv chi-squared yi Statistic p-value 4.8585 2.7548 2.8894 -1.7276 0.9877 0.9971 0.9969 0.9580 P Z L* Pm pl n ua al P statistic requires number of panels to be finite Other statistics are suitable for finite or infinite number of panels n va ll fu Appendix 20: Stationary test for first difference of money supply ratio oi m xtunitroot fisher d.m2_gdp , dfuller lag(1) trend (47 missing values generated) nh Ho: All panels contain unit roots Ha: At least one panel is stationary Number of panels = Avg number of periods = AR parameter: Panel means: Time trend: Drift term: Asymptotics: T -> Infinity at Fisher-type unit-root test for D.m2_gdp Based on augmented Dickey-Fuller tests z z vb k ADF regressions: lag 0.0000 0.0000 0.0000 0.0000 n va P statistic requires number of panels to be finite Other statistics are suitable for finite or infinite number of panels an Lu 78.5803 -6.4935 -8.2119 12.2045 om p-value l.c Statistic gm P Z L* Pm jm Inverse chi-squared(14) Inverse normal Inverse logit t(39) Modified inv chi-squared ht Panel-specific Included Included Not included 17.71 ey t re 75 Appendix 21: Stationary test for natural logarithm of GDP per capita t to xtunitroot fisher ln_gdp_per_capita , dfuller lag(1) trend (38 missing values generated) ng Fisher-type unit-root test for ln_gdp_per_capita Based on augmented Dickey-Fuller tests hi ep Number of panels = Number of periods = Ho: All panels contain unit roots Ha: At least one panel is stationary w n AR parameter: Panel means: Time trend: Drift term: lo ad Panel-specific Included Included Not included Asymptotics: T -> Infinity ADF regressions: lag ju y th yi Inverse chi-squared(14) Inverse normal Inverse logit t(39) Modified inv chi-squared 19 pl p-value 3.0405 2.8816 2.8933 -2.0712 0.9990 0.9980 0.9969 0.9808 P Z L* Pm ua al Statistic n P statistic requires number of panels to be finite Other statistics are suitable for finite or infinite number of panels n va ll fu oi m Appendix 22: Stationary test for first difference of natural logarithm of GDP per capita at nh xtunitroot fisher d.ln_gdp_per_capita , dfuller lag(1) trend (45 missing values generated) z Fisher-type unit-root test for D.ln_gdp_per_capita Based on augmented Dickey-Fuller tests z vb Number of panels = Number of periods = AR parameter: Panel means: Time trend: Drift term: Asymptotics: T -> Infinity jm ADF regressions: lag 23.1102 -1.3151 -1.4753 1.7217 0.0585 0.0942 0.0741 0.0426 n va ey t re P statistic requires number of panels to be finite Other statistics are suitable for finite or infinite number of panels an Lu p-value om Statistic l.c gm P Z L* Pm k Panel-specific Included Included Not included Inverse chi-squared(14) Inverse normal Inverse logit t(39) Modified inv chi-squared 18 ht Ho: All panels contain unit roots Ha: At least one panel is stationary 76