Productivity in Luxembourg
Chiara Peroni
Research division, STATEC
Conseil Economique et Social, 1st July 2016
The aim of this talk
I
Why productivity?
I
I
Source of economic growth
Measure of living standard
I
Productivity measures the ability to use resources
efficiently and to produce more
I
Different methods & data sources give insights into
different sources of productivity growth
I
The measurement of productivity is challenging
Outline of this talk
1.
2.
3.
4.
5.
6.
7.
8.
Definition of productivity
Drivers of productivity
The study of productivity in Luxembourg:
Measures based on aggregate data
Measures based on firm-level data
The evolution of productivity in Luxembourg
Challenges of measuring productivity
The research ahead
Productivity is..
.. not working harder but working smarter! (OECD)
I
Productivity expresses how well countries/industries/firms
use their resources.
I
Measured as ratio of output to inputs used to produce
those output.
I
It is a relative concept.
Productivity measures
I
Partial vs. total factor productivity measures:
I
I
I
Labour productivity compares output to the labour
input.
Total Factor Productivity (TFP) compares output to
the stock of capital and labour.
They are related: ∆(GDP/L) = ∆(K /L) + ∆(TFP)
Labour prod. growth = capital deepening + TFP growth
the best restaurant in the world..
Source: http://www.osteriafrancescana.it/
Drivers of productivity
What explains productivity? Why productivity differ?
I
I
Efficiency, technical progress, factor allocation.
Internal drivers:
I
I
I
I
I
I
Quality of inputs.
Intangible assets: know-how, organisation, reputation.
Knowledge capital: skills, management, HR practices.
Innovation and R&D.
More recently: workers’ incentives, job satisfaction.
External drivers:
I
Technological spillovers, trade, market structure.
Productivity and efficiency
Concepts of efficiency:
I
Productive efficiency: ability of a producer to obtain
maximal output from a given level of inputs use.
I
Allocative efficiency: state of an industry where
resources employed by most productive producers.
How we study productivity
I
I
I
Several data sources and methods to analyse productive
efficiency and allocative efficiency.
Two projects:
LuxKLEMS: National Accounts data to compute
productivity indices at national and industry level.
LuxPROD: firm-level data to investigate issues:
I
I
Productivity, mark-ups and international trade.
Resource mis-allocation and productivity slowdown in
Luxembourg.
LuxKLEMS
National Accounts framework:
I
Countries use resources to produce goods and services:
I
Stock of capital (K), Labour (L), Energy (E), Materials
(M), and purchased Services (S).
I
Use aggregate data for many countries and industries.
I
Objective: comparative indices of Total Factor
Productivity (TFP).
I
Analyse productive efficiency and technical progress.
Productive efficiency
GDP
Efficient frontier
Di
Ci
K /L
Technical progress
Efficient frontier (t + 1)
GDP
Di,t+1
Di
Ci
Ci,t+1
Efficient frontier (t)
K /L
Changes in TFP = changes in eff. + tech. progress
TFP evolution in Luxembourg
Source: Penn world tables.
TFP (1980=100)
120
DEU
100
140
160
TFP evolution: country comparison
FRA
LUX
80
BEL
1980
1985
1990
1995
2000
Source: Penn world tables.
2005
2010
2015
Technical progress and efficiency
in Luxembourg
Source: Penn world tables.
LuxKLEMS: the work ahead
Need to adapt LuxKLEMS to the new data framework:
I
New European System of National Accounts (ESA2010):
I
I
I
I
Implemented by EU countries in 2014;
Major data revisions for all years;
R&D recorded as investment expenditure.
New industry classification NACE Rev.2.
New topics
I
I
Quality of life: does well-being matter to productivity?;
Quality of environment: environmental efficiency.
Capital
GDP
Labor
WB
CO2
Productivity in Luxembourg: a summary
Productivity reflects economic cycle and structural changes:
I
Fall in GDP during crisis due to fall in TFP.
I
Fall in productivity is persistent.
I
Luxembourg productivity slow-down preceded the crisis.
I
Similar patterns across countries.
I
Aggregate conceals industry and within-industry
variations.
Studies on firm-level data:
Allocative efficiency
I
Aggregate productivity depends on efficiency in the
allocation of resources across producers.
I
State of an industry where (important portion of)
resources are employed by most productive producers.
I
Micro data inform on how industries responded to the
negative productivity shock.
I
Data: Structural Business Statistics & Business Register.
Entry, exit and size
Sources of allocative efficiency:
I
I
I
I
Some firms become more productive (within);
More productive firms become larger in size (between);
Exit of inefficient producers;
Entry of new firms.
Next: some results on labour productivity in Luxembourg
industries.
Source: Business Register data.
Sources of labour productivity growth
Transportation and Storage
1999
2002
2005
2008
2011
1999
2002
2005
2008
2011
−.2
1999
2002
2005
2008
2011
.2
0
−.05
2005
2008
2011
−.2
−.4
1999
2002
2005
2008
2011
2005
Between
Entry
Exit
−.1
−.15
−.2
2002
2002
2008
Prod. Growth
Within
0
−.05
−.1
1999
1999
Administrative Activites
.4
.05
Prof., Sci. and Tech. Activities
0
.05
Accomodation and Food Service
−.1
−.05
−.15
−.1
−.5
−.05
−.1
0
0
0
0
.05
.05
.1
Wholesale and Retail Trade
.1
Construction
.5
Manufacturing
1999
2002
Source: U. Kilinc on BR data.
2005
2008
2011
2011
.01
.005
0
.015
.005
0
.01
0
1996 1998 2000 2002 2004 2006 2008 2010
1996 1998 2000 2002 2004 2006 2008 2010
Exit Rate
Entry Rate
.01
.03
.02
.03
.02
.01
0
.015
.025
.015
0
Administrative Activites
.02
1996 1998 2000 2002 2004 2006 2008 2010
Prof., Sci. and Tech. Activities
.04
1996 1998 2000 2002 2004 2006 2008 2010
Accomodation and Food Service
.04
1996 1998 2000 2002 2004 2006 2008 2010
1996 1998 2000 2002 2004 2006 2008 2010
Transportation and Storage
.005
.01
0
.005
.01
.015
.003
.002
.001
0
Wholesale and Retail Trade
.02
Construction
.02
.004
Manufacturing
.025
Entry and exit
1996 1998 2000 2002 2004 2006 2008 2010
Source: U. Kilinc on BR data.
Some new results based on macro data:
TFP in service industries
(very preliminary)
Wholesale trade
40
95
60
TFP (1995=100)
100
105
TFP (1995=100)
80
100
120
110
Construction
2000
2005
Source: CH Di Maria on NA data.
2010
2015
2000
2005
2010
2015
Some new results based on macro data:
TFP in manufacturing industries
metals
130
80
140
90
TFP (1995=100)
100
TFP (1995=100)
150
160
110
170
120
mining
2000
2005
Source: CH Di Maria on NA data.
2010
2015
2000
2005
2010
2015
Strengths of our approach
Combines different data sources:
I
I
I
I
LuxKLEMS: productivity indices based on minimal
assumptions;
Robust to availability of new data (obviously not to
sample-wide revisions);
Evidence based on micro-data allows robustness checks
and explains aggregate outcomes;
Firm-level and national-level data give insights into
sources of productivity growth.
Challenges of measuring productivity
I
Data availability.
I
I
Inputs to production are often estimated or proxied.
I
I
I
Coverage, transmission lags, revisions;
Capital stock: cumulated investment by asset type;
Human capital highly problematic.
The measurement of productivity in services is difficult.
I
I
Output of services is hard to define and measure.
Financial services: FISIM to capture implicitly priced
services of banks.
Take-homes:
I
I
I
I
Productivity compares output to inputs to production.
Different methods give insights into sources of
productivity growth: productive efficiency; allocative
efficiency; changes in technology.
A key issue in the analysis of productivity is the definition
and measurement of inputs and output(s).
Why is this important?
I
I
Apparent slow-down in Luxembourg’s productivity.
The results have policy implications.
Further research
I
I
I
I
I
Continue the current work to produce up-to-date figures.
Continue research effort on productivity drivers.
Measures of productivity in key industries.
New important issues (well-being, environment).
We need data!
References
Kilinc, U. 2016. Resource Misallocation and Productivity Slowdown: Luxembourg’s
Manufacturing in the Post-Recession, Mimeo.
Kilinc, U. 2015. Productivity, Markups and International Trade: The Case of Small
Open Economy, Mimeo.
Ben-Aoun, L. and U. Kilinc. 2015. Dynamics of Net Job Creation, Firm Entry and
Exit in Luxembourg’s Main Sectors, in Stabilité dans un environnement à risques,
Perspectives de Politique Économique, Bilan Compétitivité 2015, no. 30, pp. 218-222.
Kilinc, U. 2014. Aspects of Business Demography, in Dynamiques des entreprises du
Luxembourg, Compétitivité de la nation, Cahier économique, No 118, pp. 64-71.
Kilinc, U. 2014. Factor Allocation and Firm-Level Productivity Dynamics in
Luxembourg’s Manufacturing Sector, Economie et Statistiques Working papers du
STATEC, No 71.
Kilinc, U. 2012. A Note on Measuring Firm-Level Capital Stock and Productivity in
Luxembourg’s Manufacturing Sector, in 2012 Competitiveness Report, Perspectives de
Politique Économique, Observatoire de la Compétitivité, pp. 165-179.
Kilinc, U. 2014. Factor Allocation Dynamics in Manufacturing Sector, in Dynamiques
des entreprises du Luxembourg, Compétitivité de la nation, Cahier économique, No
118, pp. 42-47.
References
DiMaria, C.H., Sarracino, F., and Peroni,C. 2015. Happiness matters: the role of
well-being in productivity mimeo.
DiMaria, C.H., Sarracino, F., and Peroni,C. 2015. Happiness matters: the role of
well-being in productivity in Stabilité dans un environnement à risques, Perspectives
de Politique Économique, Bilan Compétitivité 2015, no. 30: 25336.
Peroni, C. 2014. Total Factor Productivity at the industry level , in Dynamiques des
entreprises du Luxembourg, Compétitivité de la nation, Cahier économique, No 118,
pp. 46-58.
Peroni, C. 2013. Productivity and Competitiveness in Luxembourg, in 2013
Competitiveness Report, Perspectives de Politique Économique, Observatoire de la
Compétitivité, no 27: pp. 169-195.
DiMaria, C.H. and Peroni,C. 2012. A new unit labour cost changes decomposition:
four pillars of cost competitiveness recovery, mimeo.
Peroni, C. 2012. Environmental efficiency indices: towards a new approach to
green-growth accounting, Economie et Statistiques Working Paper Series, no. 61,
October 2012.
Peroni, C. 2012. Productivity and competitiveness in Luxembourg: the evolution of
total Factor productivity in luxembourg from 1995 to 2010 Perspectives de Politiques
Economiques, no. 25: 164.
Peroni, C. 2011. A review of Total Factor Productivity of Luxembourg in 2011
Competitiveness Report, Perspectives de Politique Economique, Observatoire de la
Compétitivité, 2011, NO. 17: 198212.
References
Dubrocard, A., Gomes Ferreira, Y.,and Peroni, C. 2010. Productivité et comptitivité
au Luxembourg: une comparaison par Pays et par branches Perspectives de Politiques
Economiques, no. 14: 183.
Ciccone, J, and DiMaria, C.H. 2008. Productivité et comptitivité! Perspectives de
Politiques Economiques, no.8: 154.
Asikainen, A.L., and Dubrocard, A. 2008. Innovation et Productivité Perspectives de
Politiques Economiques, no.9: 168.
Asikainen, A.L. 2008. Innovation and Productivity in Luxembourg Economie et
statistiques, no. 23: 122.
Ciccone, J, and DiMaria, C.H. 2006. Vers des mesures de la productivité totale des
facteurs au Grand-Duché de Luxembourg, défis et pistes de recherche Perspectives de
Politiques Economiques, no. 6: 11424.
Ciccone, J, and DiMaria, C.H. 2006. La productivité totale des facteurs au
Luxembourg, Cahiers Economiques, no. 102: 1-88.
Ciccone, J, and DiMaria, C.H. 2003. Analyses théoriques et empiriques des
déterminants de la productivité globale des facteurs: une application au Grand-Duché
de Luxembourg Perspectives de Politiques Economiques, no. 1: 1127.
References
Feenstra, Robert C., Robert Inklaar and Marcel P. Timmer. 2015. The Next
Generation of the Penn World Table.American Economic Review, 105(10): 3150-82.
Syverson, C. 2011. What determines productivity? Journal of Economic Literature
2011, 49:2, 326365.
Coelli, Rao, ODonnell and Battese (2005). An Introduction to Efficiency and
Productivity Analysis (2nd Ed.). Springer.