Corporate Tax Competition among County-Level Governments: A Quasi-Natural Experiment from China Yukun Sun Zhongnan University of Economics and Law June, 2019 1 / 23
Motivation ◮ County level governments have played a vital part for China’s economic growth ◮ Should we expect competion ◮ Shold we expect a “race to the bottom” ◮ Channels of tax competition 2 / 23
Literature ◮ Theory ◮ Wilson(1986, 1999) Theory of tax competition ◮ Empirical ◮ Agrawal (2015) Sales tax competition with multiple federations ◮ Liu and Martinez-Vazquez(2014) Provincial level tax competition in China ◮ Yu et al(2016) Political competition across Chinese cities ◮ In Chinese ◮ Long et al(2014); Yang and Yin(2014); Wang and Fang (2015); Xie and Fan (2015); Li(2015); 3 / 23
Mechanisms ◮ Focus on sub provincial level tax competition ◮ Government tasks are performed by prefectures and counties ◮ Most tax collectors are employed at county or lower level ◮ Lower statutory tax rate in “special economic zones” ◮ Tax holidays with complete or partial tax expemptions ◮ Tax rebates for reinvested earnings ◮ Control of level of enforcement 4 / 23
Potential Contribution ◮ Investigate county tax competition in China, and try to investigate competition among multiple federations ◮ Use varies research designs 5 / 23
A Simple Model The utility function � B ( U x ( xq ( b ) , P )) db + U l ( g l ) + U c ( g c ) U = (1) 0 The gross of tax price 1 + τ l b ∈ 0 , [ min ( B l , B c )] q ( b ) = 1 + τ c b ∈ [ min ( B l , B c ) , 1] 1 + τ l + τ s b ∈ [ B l , B c ] ifB l > B c 6 / 23
A Simple Model The indirect utility function � B ( U x ( xq ( b ) , P )) db + U l ( g l ( τ l , τ b , B c , B l ))+ U c ( g c ( τ l , τ b , B c , B l V = 0 (2) Local government objective function � B F l [( τ l , τ b , B c , B l )] = ( U x ( xq ( b ) , P )) db + U l ( g l ( τ l , τ b , B c , B l )) 0 (3) Upper level government objective function � B F c [( τ l , τ b , B c , B l )] = ( U x ( xq ( b ) , P )) db + U c ( g c ( τ l , τ b , B c , B ) )+ αU 0 (4) 7 / 23
A Simple Model Impact of tax rates and tax bases ∂x ( b ) 1 dg j = τ j [ x z + τ l ( b ls + b j ) X j x ( b )] (5) dB j ∂P dg i ∂x ( b ) 1 ∂x ( b ) 1 = τ j τ i [ Ex z x ( b ) + τ l ( b ls + b j ) X j x ( b )] (6) dB j ∂q ( b ) ∂P dg j = X j [1 + τ j ( ∂x ( b ) 1 ∂x ( b ) 1 x ( b ) + b j + b ls x ( b ))] (7) dτ j ∂q ( b ) ∂P dg i b ls x ls ∂x ( b ) x ( b ) + ( b s + b ls ) ∂x ( b ) 1 1 = τ i X i [ x ( b )] (8) b ls + b i ∂q ( b ) dτ j x i ∂P 8 / 23
Propositions Proposition: Given certain conditions, if governments independently choose their tax bases, tax competition lead to local governments tax the entire base 9 / 23
Data and Measurement Data: Survey of industiral firms; Prefecture and county level data Measure of tax incentives: average of effective tax rate Firm level AETR define as CIT paid devided by pre-tax profits Then calculate the AETR for all firms in a given region 10 / 23
Trends in Tax Rates—Average Effective Tax Rate (.3193758,1] (.271371,.3193758] (.2208769,.271371] [.0013423,.2208769] 11 / 23
Trends in Tax Rates—Weighted Effective Tax Rate (.325509,1] (.2647329,.325509] (.1782587,.2647329] [.001045,.1782587] 12 / 23
Spatial Econometrics ◮ Model setup y it = λ Σ n j =1 W ij y jt + βX it + υ i + η t + ǫ it (9) ◮ Firm level data were obtained from the Annual Survey of Industrial Firms 13 / 23
Spatial Econometrics Results Table: Spatial Econometrics AETR WAETR Prefecture λ 0.1037*** 0.1306*** 0.076* (0.0211) (0.0319) (0.042) FDI share -0.062*** -0.071*** (0.017) (0.019) County FE Y Y Y Year FE Y Y Y Observations 15,311 15,311 1023 14 / 23
Domestic and Foreign Table: Domestic and Foreign Domestic Foreign 0.0823*** 0.2917*** λ (0.0191) (0.0632) County FE Y Y Year FE Y Y Observations 5,621 5,621 R 2 0.652 0.509 15 / 23
Turning Counties into Urban Districts ◮ Facilitated by the 1993 State Council approval of a Ministry of Civil Affairs report that established standards for Conversion of counties into urban districts ◮ Prefecture: Coordinate development throughout its entire territory ◮ Counties: Loss of any substantial power to control itself 16 / 23
Turning Counties into Urban Districts Table: Changes of Administrative Divisions Year County to Division City to Division 1998 10 5 1999 10 8 2000 33 24 2001 20 6 2002 23 9 2003 9 6 2004 5 1 2005 0 0 2006 7 0 17 / 23
Province 18 / 23
Prefecture 19 / 23
Differences-in-Differences Regression equation AETR it = α + β 1 D i + β 2 D t + β 3 D i × D t + γ X + ǫ it (10) 20 / 23
Difference-in-Differences Table: Difference-in-Differences ETR 0.0068*** β 3 (0.0021) -0.0186*** ROA (0.0052) -0.0279*** Debt (0.0019) lnpop 0.0826*** (0.0103) gdpper 0.0110*** (0.00374) Observations 15,138 R 2 0.178 21 / 23
Conclusions ◮ Local governments do compete ◮ Does it matter? What are the consequences? 22 / 23
To Do List ◮ AETR for several consecutive years ◮ Add vertical competition in the mix ◮ Channels: Province-Prefecture-Country or Province-County ◮ Premutation tests ◮ Different spatial weighted matrix 23 / 23
Recommend
More recommend
