Business Cycle Accounting - China vs India

Business Cycle Accounting: China vs. India Christer Ljungwall* and Gao Xu** *China Center for Economic Research, CCER, and the Stockholm School of Economics. ** World Bank Office, Beijing, China (EASPR). The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the view of the World Bank, its Executive Directors, or the countries they represent. Outline      Motivation Literature Review Quantitative Method (Business Cycle Accounting)  Benchmark Model  Accounting Procedures Empirical Findings  Comparison of Business Cycle Facts  Comparison of Business Cycle Accounting Results Conclusions July 3, 2008 2 Motivation  The rise of China and India is one of the most significant economic developments in nowadays    Great achievement of human development: better life for 1/3 of world population Major representatives of emerging markets: BRIC Big players in the global market: made-in-China, service outsourcing into India  Similar development path of China and India   Start point: planned economy of China and rigid state control in India Market oriented reform and integration into the world economy July 3, 2008 3 Motivation  Differences between China and India    Average growth: 9.8% of China vs. 3.4% of India Investment-led growth pattern of China vs. consumption-led growth in India Strong competitiveness of China’s manufacturing sector vs. advantages of service outsourcing in India  It is of both theoretic and practical interest to compare China and India within a rigorous quantitative framework July 3, 2008 4 Literature Review  Business Cycle Accounting (BCA)      Proposed by Mulligan(2002) and Chari, Kehoe and McGratten (2007) A DSGE model with time varying wedges Measure the wedges so the model replicates data exactly. Inspect measured wedges to analysis shocks. Application: Chakraborty (2004), Kobayashi and Inaba (2006) to investigate Japan’s recession. Lama (2005) used it to identify business cycle sources for Argentina, Brazil and Mexico. Cavalcanti (2004) July 3, 2008 5 Literature Review  Business cycle research on Chinese economy   Descriptive study and summary statistic calculations: Qian (2004), Lu and Qi (2006), Liu (2006) SVAR: Zhang and Wan (2005), Xu (2007)  Little quantitative research on Indian business cycle fluctuation July 3, 2008 6 Benchmark Model  Model setup  max   t t ( s t )U (ct ( s t ), lt ( s t )) ct ,lt , xt t  0 st Household  s.t. ct ( s t )  [1   xt ( s t )]xt ( s t )  [1   lt ( s t )]wt ( s t )lt ( s t )  rt ( s t )kt ( s t 1 )  tt ( s t ) (1   )kt 1 ( s t )  (1   ) kt ( s t 1 )  xt ( s t )  Firm {kt ( s ),lt ( s )} max t At ( s t ) F (kt ( s t 1 ), lt ( s t ))  rt ( s t )kt ( s t 1 )  wt ( s t )lt ( s t ) t ct ( s t )  xt ( s t )  g t ( s t )  yt ( s t )  Resource constraint July 3, 2008 7 Benchmark Model  Four wedges (shocks)  At ( s t ) Efficiency wedge  frictions which cause inputs to be used inefficiently financial frictions, etc. sticky wage, powerful labor union, etc. government consumption, net exports  investment wedge 1/[1   xt ( s t )]  [0,1]   1   lt ( s t )  [0,1] labor wedge   gt ( s t ) government (exogenous demand) wedge  July 3, 2008 8 Benchmark Model  Equilibrium conditions   Optimal condition for consumption-leisure choice  U lt ( s t ) U ct ( s t )  [1   lt ( s t )] At ( s t ) Flt ( s t ) Euler equation U ct ( s t )(1   )[1   xt ( s t )] st 1     ( s t 1 | s t )U ct 1  At 1 ( s t 1 ) Fkt 1 ( s t 1 )  (1   )[1   xt 1 ( s t 1 )]  Production technology yt ( s t )  At ( s t ) F (kt ( s t 1 ), lt ( s t ))  Resource constraint ct ( s t )  xt ( s t )  g t ( s t )  yt ( s t ) July 3, 2008 9 Accounting Procedures  Shocks    st  [at  xt  lt st 1  Hst   t 1  gt ], st 1 ~ i.i.d. N (0,V ), V  QQ  Estimation of wedges      Log-linearize equilibrium conditions Solve linearized system with Blanchard-Kahn (1980) method Use Kalman filter to write likelihood function Estimate parameters with MLE combined with prior from long-run relationships (Bayesian approach) Estimate wedges with Kalman smoothing algorithm July 3, 2008 10 Accounting Procedures  Counterfactual experiments   Marginal effect of each wedge: let one wedge fluctuate and keep the rest fixed, and simulate. Effect of a combination of wedges: let a subset of wedges fluctuate and keep the rest fixed, and simulate. Annual GDP by expenditure data, deflate nominal variables with GDP deflator to construct real series. 4 observation series (to avoid singularity problem): GDP, private consumption, investment, government consumption plus net exports (all are log deviations from their HP trends). 11  Data   July 3, 2008 Comparison of Business cycle Facts  Growth rate   China: high and persistent India: relatively low Real GDP Growth 20 15 Real GDP growth (%) 10 5 0 -5 China -10 India 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 July 3, 2008 2006 12 Comparison of Business cycle Facts  Expenditure structure    China: significant role of investment India: predominant role of private consumption Similarity: Shrinking relative size of private consumption China: Real GDP Growth and GDP Decomposition 16 14 Different Expenditure Structure in 2006 Unit: percent of GDP China India Private Consumption Investment Government Consumption 1/ 1/ Net export is included. India: Real GDP growth and GDP decomposition 90 80 9 8 7 6 5 4 3 2 1 0 -1 -2 36 43 21 59 32 9 60 50 Share in GDP (%) 70 60 50 40 30 20 10 0 Share in GDP (%) 12 40 30 20 10 0 10 8 6 4 2 0 Real grwoth (%) 1980/81 1982/83 1984/85 1986/87 1988/89 1990/91 1992/93 1994/95 1996/97 1998/99 2000/01 2002/03 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 Real GDP growth (RHS) Investment Private consumption Government consumption plus net exports Real GDP growth (RHS) Investment Private consumption Government consumption plus net exports July 3, 2008 13 2004/05 Real growth (%) Comparison of Business cycle Facts  Business cycle fluctuation   China: big output fluctuation, even bigger consumption volatility, high output persistence, investment lag output India: small output fluctuation, smaller consumption volatility, less output persistence, investment co-moves with output A. China: 1978-2006 variables Output Private Consumption Investment Government Consumption 2/ B. India: 1980/81-2005/06 variables Output Private Consumption Investment Government Consumption 2/ SD% 1.70 1.43 7.38 7.80 Cross Correlation of Output with variable at Lag k= -2 .01 - .09 .12 - .45 -1 .32 .02 .53 - .11 0 1.00 .70 .64 .00 1 .32 .20 .13 .41 2 .01 .10 - .14 .09 SD% 3.20 4.13 8.02 8.98 Cross Correlation of Output with variable at Lag k= -2 .17 - .09 - .48 - .10 -1 .72 .28 - .17 - .18 0 1.00 .65 .27 - .39 1 .72 .68 .65 - .33 2 .17 .43 .80 .13 July 3, 2008 14 Comparison of BCA Results China: Data and Predictions of Models with Just One Wedge: Output (1978-2006) 15 Percentage Deviation from Trend 10 5 0 -5 -10 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 15 arial Observed fluctuations of real GDP Model with Investment Wedge Model with Government Consumption Wedge Model with Efficiency Wedge Model with Labor Wedge July 3, 2008 Comparison of BCA Results Indian: Data and Predictions of Models with Just One Wedge: Output (1980/81-2005/06) 10 8 6 4 2 0 -2 -4 -6 -8 -10 -12 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 Percentage Deviation from Trend Observed fluctuations of real GDP Model with Investment Wedge Model with Government Consumption Wedge Model with Efficiency Wedge Model with Labor Wedge July 3, 2008 16 Comparison of BCA Results  Main driving force in both countries: efficiency wedge (Solow residual)  Important role played by technology advances and infrastructure change Missing factors of RBC model   Bigger damping effect of labor wedge points to more labor market rigidities in India Ignorable roles played by financial frictions and government consumption in both countries  July 3, 2008 17 Conclusions  Both of China and India’s business cycle fluctuations are mainly driven by efficiency wedge (including technology advance and institutional change)  More rigid labor market (sticky wage and powerful labor union) is spotted in India Minor roles played by financial frictions and government consumption in both countries  July 3, 2008 18 Thanks! comments welcome July 3, 2008 19