Decomposing Factor Income Distributions: The Role of Trade, - - PowerPoint PPT Presentation

Decomposing Factor Income Distributions: The Role of Trade, Technology and Consumption Marcel Timmera Bart Losa Robert Stehrerb Gaaitzen de Vriesa

(a) Groningen Growth and Development Centre, University of Groningen (b) The Vienna Institute for International Economic Studies (WIIW)

Presentation at final WIOD conference, Groningen, April 24, 2012

This project is funded by the European Commission, Research Directorate General as part

• f the 7th Framework Programme, Theme 8: Socio-Economic Sciences and Humanities.

Grant Agreement no: 225 281

Background

• Increasing income inequality within countries in past decades,

both in advanced and emerging economies (Autor, 2010; Goldberg and Pavcnik, 2007)

• Various determinants have been suggested both in theory and

empirics (see key-note Adrian Wood):

• International trade
• Skill biased technological change (SBTC)
• Conclusion of older literature (cross-country empirics) : SBTC is

dominant explanation (e.g. Michaels et al. 2010)

• New trade channels: Increasing global production sharing

(Feenstra and Hanson, 1999; Costinot, Vogel and Wang 2012)

• Changes in patterns of consumption (Yi and Zhang, 2010)

Background

• Contribution of this paper: a new ex post decomposition analysis
• f income inequality within countries relying on an input-output

modelling of world production, including trade in intermediate goods and services

• Decompose wage income inequality within 40 countries across

workers with different skills into:

• Changes in relative wages
• Changes in labour quantities needed to produce products

(includes off-shoring of activities and technological change in global production)

• Changes in final demand for products
• New method inspired by input-output decomposition techniques

from Leontief (1949) in a multi-regional setting (related to recent work by e.g. Yi, Johnson and Bems, 2011, 2012)

Methodology

• Define product as final good produced in industry i in country j
• All activities needed to produce a product (i,j) is denoted by shorthand

global production of (i,j)

• Workers are of different skill classes (s)
• Decompose the change in income of worker s in country j over period

1995-2008 into:

• Change in wage for (s,j)
• Change in number of workers (s,j) needed in global production of

product (i,x)

• Change in share of country j in total s needed for (i,x)

(depends on trade in activities; if x=j this is “off-shoring”)

• Change in s needed for production of (i,x)

(technical change in global production)

• Changes in final demand for product (i,x)

• World input output tables (from WIOD) for global production

structure (A) and final demand (ef)

• Quantities of labour used by skill (by industry and country, l)

(WIOD)

• Wages of labour used by skill (by industry and country, w)

(WIOD)

• Efficiency levels of labour of a skill type across countries (π,

GGDC productivity level database)

Methodology and Data

Example: Global production of car from German transport industry

DEU DEU Other EU Other EU ROW ROW CHN CHN JPN JPN KOR KOR USA USA

20 40 60 80 100 1995 2006 Shares (%)

Income earned in global production of German car (1995 and 2006)

Example: Global production of car from German transport industry

Own Low skill labour Own Low skill labour Own Medium skill labour Own Medium skill labour Own High skill labour Own High skill labour ROW Low skill labour ROW Low skill labour ROW Med skill labour ROW Med skill labour ROW High skill labour ROW High skill labour

20 40 60 80 100 1995 2006 Shares (%)

Income earned in global production of German car (1995 and 2006)

Example: Global production of car from German transport industry

• Decomposition of income of low skilled workers in various

countries involved in the global production of final output from the German transport equipment industry (ratio of 2008 over 1995 in current $)

Total wage income Wage per worker Share in workers needed Workers needed

DEU 0.58 1.67 0.69 0.50 CZE 3.06 3.90 1.57 0.50 HUN 2.41 1.83 2.63 0.50 POL 1.55 1.59 1.95 0.50 SVK 1.42 0.57 4.98 0.50 Low-skilled

Example: Global production of car from German transport industry

• Decomposition of income of low skilled workers involved in the

global production of final output from the German transport equipment industry (ratio of 2008 over 1995 in current $)

Total wage income Wage per worker Share in workers needed Workers needed Total wage income Wage per worker Share in workers needed Workers needed

DEU 0.58 1.67 0.69 0.50 1.01 1.99 0.70 0.72 CZE 3.06 3.90 1.57 0.50 5.09 3.83 1.84 0.72 HUN 2.41 1.83 2.63 0.50 5.33 2.51 2.95 0.72 POL 1.55 1.59 1.95 0.50 5.38 1.53 4.87 0.72 SVK 1.42 0.57 4.98 0.50 4.98 0.73 9.44 0.72 Low-skilled High-skilled

More examples

Total wage income Wage per worker Share in workers needed Workers needed Total wage income Wage per worker Share in workers needed Workers needed

China electronics China 0.47 2.81 0.87 0.19 1.70 3.56 0.59 0.81 Japan 0.41 1.11 1.92 0.19 1.15 1.14 1.25 0.81 Korea 0.44 1.15 2.01 0.19 2.39 1.30 2.25 0.81 US transport US 0.55 1.44 0.75 0.51 1.12 1.62 0.95 0.73 Mex 1.11 2.32 0.93 0.51 1.62 2.38 0.93 0.73 Low-skilled High-skilled

Conclusions and Next steps

• Preliminary conclusions:
• Decomposition is feasible and has intuitive interpretations
• New measure of biased technological change in a global

production system setting

• Find strong evidence of skill-bias in global production
• Next steps:
• Make decompositions for global production of all products
• Add in the changes in demand for various products (e.g.

increasing demand for durables; increasing demand for final products from China)

• Fully decompose the change in labour income across

workers in each country