china: input-output tables 1992-2010 - university...
TRANSCRIPT
China: Input-Output Tables 1992-2010
Douglas Nyhus
Inforum, University of Maryland
20th Inforum World Conference
Florence, Italy
Why?
• Data base for MUDAN model
• Check consistency of Chinese data
• Examine structural change
• Economic imbalances?
• Energy use
• CO2 generation
Inforum World Conference September 2012 2
Basic Data Sources
• Input-output tables: 1992, 1997, 2002, 2007
• China Statistical Yearbook (CSY): 1992-2011
• CSY data in electronic format from 1995-2011
• China Labor Statistics Yearbook: 2001-2010
• China Economic Census Statistics 2004, 1995
• Bilateral Trade model data base
• National Bureau of Statistics of China (monthly data)
Inforum World Conference September 2012 3
Major Problem Areas
• Changes in Statistical System: NMP (1992) to SNA(1997,2002,2007)
• Revisions in statistical coding 1995,2004,2007
• Major changes in scale—GDP revised up 10% in 2009 without revisions in production data
• Tables (CSY) appear and disappear
• Labor (employment) data weak
• Factor income data at national level is nonexistent
Inforum World Conference September 2012 4
Major Strengths
• Historical tables relatively consistent
• Gradually improving data base – especially for services
• Agriculture and Industry data (Mining, Manufacturing and Utilities) are consistent enough over time to be useable
• Expanding detail from the national accounts
Inforum World Conference September 2012 5
Plan I
• Balance tables in current prices
• Balance full four quadrant IO tables at one time—intermediate (59) sectors, final demands and value added (wages, depreciation, taxes less subsidies, surplus)
• Use at much data as can be gleaned from the CSY and other data sources—methods vary greatly due to coverage, general appropriateness, etc.
Inforum World Conference September 2012 6
Plan II
• For each cell in the large matrix attempt to construct a consistent time series between IO table years—e.g. 1992-1997, 1997-2002, etc
• Then “benchmark” the end points based on the IO table and generate a “fix” file for starting points for the intervening years
• This is applied for all gross output of all sectors
Inforum World Conference September 2012 7
Plan III
• Balance backwards and forwards—starting points are important—e.g. balance 1992-96 going forward and the balance 1997-93 going backwards – Remember interdyme can operate backwards as well as forwards in time.
• Use estimates from both of the estimates as a starting point for final balance
Inforum World Conference September 2012 8
Plan IV
• After balancing table we then balance bridge matrices for investment –52 investing industries (columns) and 59 rows
• Similar household consumption matrices—urban (59 rows, 24 columns), rural (59 rows,10 columns)
Inforum World Conference September 2012 9
Plan V
• Deflate the table to obtain constant price estimates
• “Known” prices: (1) gross output; (2) import prices (3) export prices (4) rural consumption prices and (5) urban consumption prices
• Implicit prices for each row of the table are implied—(1) domestically produced goods for particular uses are necessarily different from production prices—e.g. Urban and rural households; (2) user prices (mixture of domestic and import prices).
• “User” prices are used for deflation
Inforum World Conference September 2012 10
Plan VI
• Check estimate of deflated GDP with NA
• How different? Why?
• Estimate service industry output prices– value added prices are the only ones published and these only implicitly
• Assume the implicitly published prices are constructed using a double deflation method
Inforum World Conference September 2012 11
Plan VII
• Estimate new service industry “output” prices that are consistent with “deflated” valued added and the in the IO Table
• Use the resulting matrices as the basis for estimating MUDAN
• Reconcile 2002 estimates for GDP with National Accounts
Inforum World Conference September 2012 12
Strengths of Plan
• Internal consistency
• Ability to examine any series as a time series and not just a point estimate
• Ease of updating – documentation –all preparation work is done starting with excel tables in the CSY followed by G7 add files (scripts) preparing the work for interdyme which integrates and completes the work.
Inforum World Conference September 2012 13
Weaknesses
• Estimates of GDP in 2002 prices not consistent with National Accounts
• Only total household consumption (rural and urban) is available in constant price units on product side national accounts
• Some estimates of the implied prices of domestic production for domestic consumption are not reasonable—indicating a defect in data or method
Inforum World Conference September 2012 14
Some Results
Inforum World Conference September 2012 15
Inforum World Conference September 2012 16
Real GDP Real GDP VAD pr i ces vs " Out put " pr i ces
330790
189473
48156
1995 2000 2005 2010
gdpR vadgdpR
Industry and VAD Price GDP
CSY GDP
Inforum World Conference September 2012 17
GDP and Component s GDP and Component s 2002 = 1.
3. 53
1. 90
0. 27
1995 2000 2005 2010
l oggdpR l ogcr Rsum l ogcuRsum l ogi f aRsum l ogexRsum l ogi mRsum
Note: Log scale
Rural HH
Exports
GDP
Imports
Inv
Inforum World Conference September 2012 18
Def l at or s Def l at or s 2002 = 1.
1. 56
1. 02
0. 48
1995 2000 2005 2010
gdpD cr D cuD i f aD exD i mD
GDP
Investment
Rural HH
Urban HH
Exp&Imp
Inforum World Conference September 2012 19
1 Far mi ng 1 Far mi ng i n 2002 pr i ces
46975. 63
23537. 32
99. 01
1995 2000 2005 2010
l out 1 l ddem l ex1 l i m1 l hhc l ddemgdp
Note: Log Scale GDP = dom dem in 2002
Output & Domestic dem
Households
Imports
Exports
Missing Or wrong data
Inforum World Conference September 2012 20
39 El ect r oni c and communi cat i on equi pment 39 El ect r oni c and communi cat i on equi pment i n 2002 pr i ces
68392. 83
34400. 55
408. 27
1995 2000 2005 2010
l out 39 l ddem l ex39 l i m39 l hhc l ddemgdp
Output
households
Imports
Exports
GDP
Inforum World Conference September 2012 21
56 Heal t h car e, spor t s and soci al wel f ar e 56 Heal t h car e, spor t s and soci al wel f ar e pr i ces of out put , expor t s, i mpor t s, domest i c use
2. 28
1. 30
0. 33
1995 2000 2005 2010
pr i ces56 pr i ceexp56 pr i cei mp56 pr i cecu56 pr i cecr 56 gdpD
Output
GDP
Urban HH
Rural HH
PRICES
Inforum World Conference September 2012 22
39 El ect r oni c and communi cat i on equi pment 39 El ect r oni c and communi cat i on equi pment pr i ces of out put , expor t s, i mpor t s, domest i c use
1. 45
1. 01
0. 56
1995 2000 2005 2010
pr i ces39 pr i ceexp39 pr i cei mp39 pr i cecu39 pr i cecr 39 gdpD
GDP
Rural HH
Urban HH
Output
Exp&Imp
Thank You!
Inforum World Conference September 2012 23