Industry & State Level Value Added and Productivity - PowerPoint PPT Presentation

Industry & State Level Value Added and Productivity Decompositions Shipei Zeng 1 , Stephanie Parsons 1,3 , W. Erwin Diewert 1,2 , and Kevin J. Fox 1 1 School of Economics, UNSW Sydney 2 Vancouver School of Economics, University of British Columbia 3 Reserve Bank of Australia November 29, 2018 Shipei Zeng & Stephanie Parsons Value added decomposition November 29, 2018 1 / 33

Identifying changes in value added Value added decomposition ◮ Parametric and non-parametric estimation of production frontiers ◮ A recent decomposition proposed by Diewert and Fox (2017) • Free Disposal Hull (FDH) and index number theory • Rule out technical regress • A non-parametric approach involving only observable data Shipei Zeng & Stephanie Parsons Value added decomposition November 29, 2018 2 / 33

Identifying changes in value added Value added decomposition ◮ Value added decomposition for Australian market sector industries • 12 selected industries and 16 market sector industries • Decomposition at an aggregate level and an industry level • Sectoral explanations for Australian TFP change ◮ Simple enough to be implemented by national statistical offices • Data cubes from Australian Bureau of Statistics • R package: dfvad Shipei Zeng & Stephanie Parsons Value added decomposition November 29, 2018 3 / 33

Methodology Defining the optimal output value ◮ Cost constrained value added function R t ( p , w , x ) = max y , z { p · y : ( y , z ) ∈ S t ; w · z � w · x } ◮ Unit cost function � w · x s � c t ( w , p ) = min p · y s : s = 1 , · · · , t s Shipei Zeng & Stephanie Parsons Value added decomposition November 29, 2018 4 / 33

Methodology Defining the optimal output value ◮ Rewrite the cost constrained value added function p · y s w · x � � R t ( p , w , x ) = max w · x s : s = 1 , · · · , t s w · x = c t ( w , p ) ◮ A sequential approach which rules out technical regress Shipei Zeng & Stephanie Parsons Value added decomposition November 29, 2018 5 / 33

Methodology Explanatory factors ◮ Net output price indexes R s ( p t , w , x ) α ( p t − 1 , p t , w , x , s ) = R s ( p t − 1 , w , x ) ◮ Input quantity indexes w · x t β ( x t − 1 , x t , w ) = w · x t − 1 Shipei Zeng & Stephanie Parsons Value added decomposition November 29, 2018 6 / 33

Methodology Explanatory factors ◮ Input mix indexes R s ( p , w t , x ) γ ( w t − 1 , w t , p , x , s ) = R s ( p , w t − 1 , x ) ◮ Returns to scale δ ( x t − 1 , x t , p , w , s ) = R s ( p , w , x t ) / R s ( p , w , x t − 1 ) w · x t / w · x t − 1 = 1 Shipei Zeng & Stephanie Parsons Value added decomposition November 29, 2018 7 / 33

Methodology Explanatory factors ◮ Growth in value added efficiency p t · y t e t = R t ( p t , w t , x t ) � 1 e t ε t = e t − 1 ◮ Technical progress R t ( p , w , x ) τ ( t − 1 , t , p , w , x ) = R t − 1 ( p , w , x ) Shipei Zeng & Stephanie Parsons Value added decomposition November 29, 2018 8 / 33

Methodology Straightforward decomposition ◮ Value added growth decomposition p t · y t p t − 1 · y t − 1 = α t · β t · γ t · ε t · τ t ◮ TFP growth decomposition TFPG t = p t · y t / p t − 1 · y t − 1 α t · β t = γ t · ε t · τ t Shipei Zeng & Stephanie Parsons Value added decomposition November 29, 2018 9 / 33

Methodology A weighted average industry approach ◮ T¨ ornqvist explanatory factors: λ ∈ { α, β, γ, ε, τ } K 1 ln λ t • = 2( s kt + s k , t − 1 ) ln λ kt � k =1 ◮ Approximation of value relatives � v t � v kt K � � 1 s kt + s k , t − 1 � � � ln ≈ ln v t − 1 v k , t − 1 2 k =1 K 1 � s kt + s k , t − 1 � � α kt β kt γ kt ε kt τ kt � � = ln 2 k =1 = ln α t • + ln β t • + ln γ t • + ln ε t • + ln τ t • Shipei Zeng & Stephanie Parsons Value added decomposition November 29, 2018 10 / 33

Methodology Establishing a benchmark ◮ t = 1 A 1 = 1 , B 1 = 1 , C 1 = 1 , E 1 = 1 , T 1 = 1 ◮ t > 1 A t = α t A t − 1 , B t = β t B t − 1 , C t = γ t C t − 1 E t = ε t E t − 1 , T t = τ t T t − 1 ◮ Level value of productivity p t · y t TFP t = p 1 · y 1 · A t · B t = C t E t T t Shipei Zeng & Stephanie Parsons Value added decomposition November 29, 2018 11 / 33

Data Australian market sector ◮ 16 industries with productivity data available 1994/95-2016/17 (July-June years) ◮ 12 industries with productivity data available 1989/90-2016/17 (July-June years) ◮ Concerns about measurement problems and research periods Shipei Zeng & Stephanie Parsons Value added decomposition November 29, 2018 12 / 33

Data Australian market sector Table 1: Industry classification of the market sector in Australia Division Industry A Agriculture, Forestry and Fishing B Mining C Manufacturing D Electricity, Gas, Water and Waste Services E Construction F Wholesale Trade G Retail Trade H Accommodation and Food Services I Transport, Postal and Warehousing J Information, Media and Telecommunications K Financial and Insurance Services L Rental, Hiring and Real Estate Services M Professional, Scientific and Technical Services N Administrative and Support Services R Arts and Recreation Services S Other Services Shipei Zeng & Stephanie Parsons Value added decomposition November 29, 2018 13 / 33

Industry-level decomposition Productivity of 12 selected industries 0.25 0.20 0.15 Log Index 0.10 ABS (12) DF Method with ABS Aggregates 0.05 Weighted Industry Aggregation Method 0.00 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 Shipei Zeng & Stephanie Parsons Value added decomposition November 29, 2018 14 / 33

Industry-level decomposition DF Method with ABS aggregates 0.35 lnT lnE 0.30 lnC lnTFP 0.25 0.20 0.15 Log Index 0.10 0.05 0.00 -0.05 -0.10 -0.15 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 Shipei Zeng & Stephanie Parsons Value added decomposition November 29, 2018 15 / 33

Industry-level decomposition Weighted industry aggregation method 0.35 lnT lnE 0.30 lnC lnTFP 0.25 0.20 0.15 Log Index 0.10 0.05 0.00 -0.05 -0.10 -0.15 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 Shipei Zeng & Stephanie Parsons Value added decomposition November 29, 2018 16 / 33

Industry-level decomposition Divisions A–D Agriculture Mining 0.8 0.8 0.4 0.4 Log Index Log Index 0.0 0.0 −0.4 −0.4 1990 2000 2010 1990 2000 2010 Year Year Manufacturing Utilities 0.8 0.8 0.4 0.4 Log Index Log Index 0.0 0.0 −0.4 −0.4 1990 2000 2010 1990 2000 2010 Year Year lnTFP lnT lnE lnC Shipei Zeng & Stephanie Parsons Value added decomposition November 29, 2018 17 / 33

Industry-level decomposition Divisions E–H Construction Wholesale 0.8 0.8 0.4 0.4 Log Index Log Index 0.0 0.0 −0.4 −0.4 1990 2000 2010 1990 2000 2010 Year Year Retail Accommodation 0.8 0.8 0.4 0.4 Log Index Log Index 0.0 0.0 −0.4 −0.4 1990 2000 2010 1990 2000 2010 Year Year lnTFP lnT lnE lnC Shipei Zeng & Stephanie Parsons Value added decomposition November 29, 2018 18 / 33

Industry-level decomposition Divisions I–K, R Transport Information 0.8 0.8 0.4 0.4 Log Index Log Index 0.0 0.0 −0.4 −0.4 1990 2000 2010 1990 2000 2010 Year Year Financial Arts 0.8 0.8 0.4 0.4 Log Index Log Index 0.0 0.0 −0.4 −0.4 1990 2000 2010 1990 2000 2010 Year Year lnTFP lnT lnE lnC Shipei Zeng & Stephanie Parsons Value added decomposition November 29, 2018 19 / 33

Industry-level decomposition Thoughts on the industry results ◮ Only 4 industries showed considerable technical progress beyond 2004 • Agriculture, forestry and fishing • Retail trade • Wholesale trade • Financial and insurance services ◮ Some industries showed little technical progress even earlier than the 2004 peak • Mining (1996) • Electricity, gas, water and waste services (1998) • Information, media and telecommunications (1999) • Arts and recreation services (1991) Shipei Zeng & Stephanie Parsons Value added decomposition November 29, 2018 20 / 33

Industry-level decomposition Thoughts on the industry results ◮ The amount of inefficiency for some industries was huge • Manufacturing • Mining • Electricity, gas, water and waste services • Accommodation and food services • Arts and recreation services ◮ Some of this inefficiency is probably real and some of it probably indicates mismeasurement of inputs and outputs Shipei Zeng & Stephanie Parsons Value added decomposition November 29, 2018 21 / 33

Industry-level decomposition Industry contribution: value added shares 1.00 0.75 Agriculture Mining Manufacturing Utilities Share Construction 0.50 Wholesale Retail Accommodation Transport Information Financial 0.25 Arts 0.00 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 Year Shipei Zeng & Stephanie Parsons Value added decomposition November 29, 2018 22 / 33