OECDNBS Workshop 610 December 2004

1
OECD-NBS Workshop6-10 December 2004
Challenges for Productivity Measurement in OECD
Countries Paul Schreyer, OECD
2
Contents of presentation

• Why is productivity important?
• Concepts labour and multi-factor productivity
• Measuring output
• Measuring labour input
• Measuring capital input
• Interpreting productivity numbers
• OECD productivity database
• Conclusions

3
1. Why is productivity important?

• Basis for improvements in real incomes and
  economic well-being
• Productivity is not everything but in the long
  run it is almost everything (Paul Krugman)
• Monetary policy (inflationary pressures)
• Fiscal policy (financing of health, education,
  welfare)
• Slow productivity growth conflicting demands
  for distribution of income more likely

4
2. Concepts labour productivity

• Every measure of productivity is a ratio between
  output and input(s)
• Simplest and most frequently-encountered measure
  labour productivity

• Indicates how efficiently labour is used in
  production
• Not necessarily an indicator of effort per worker

5
Table of productivity measures
6
2. Concepts multi-factor productivity

• Ratio of output and combined capital and labour
  input

• Indicates how efficiently combined labour and
  capital are used in production
• When measured as rate of change

Cost share of capital
Cost share of labour
change in output
change in labour input
change in capital input
change in MFP
7
MFP and labour productivity

• Useful tool decomposition of labour
  productivity

change labour productivity
Effect of capital intensity
Effect of MFP
8
MFP and labour productivity

• Decomposition of labour productivity important
• Is labour productivity driven by investment in
  traditional capital (--gt rise in capital
  intensity)?
• Is labour productivity driven by investment in
  intangible capital, i.e., innovation,
  organisational change, RD (--gt change in MFP)
• Example Canada

9
Decomposition of labour productivity, Canada
10
3. Measuring output

• Good output measures required good price indices
  for deflation
• Independence of measures of output from measures
  of input important but not always observed
• Most prominent case non-market producers
• Input-based extrapolation of value-added in
  services not infrequent
• Quality change
• Well-known example ICT capital goods

11
4. Measuring labour input

• Ideal measure total number of hours worked,
  broken down by type of labour input
• Number of persons employed or number of jobs can
  generate biased measures of productivity if hours
  per person change or if there are multiple job
  holdings
• Total hours should comprise the hours of
  employees, the hours of self-employed and the
  hours of unpaid workers (e.g. family members in
  agriculture)

12
Example impact of different measures of labour
input on labour productivity, France
13
Labour input statistical sources

• 2 main statistical sources in OECD countries
• Labour force surveys (LFS)
• Establishment or firm-based surveys (ES)
• LFS
• Conducted from socio-economic perspective to
  provide reliable information on labour force
• Typically full coverage of economy
• But industry allocation not very reliable
• Also people tend to overstate their working
  hours

14
Labour input statistical sources

• ES
• Reliable employment data by industry
• Correspondence between inputs and outputs
• Only paid employment
• Not always full coverage (cut-off point for
  small-scale establishments)

LFS overstatement of hours worked
likely ES understatement of hours worked
likely Some countries have developed labour
accounts to combine sources Netherlands, and
progress also in China
15
5. Measuring capital input

• Conceptually correct measure capital services
• Capital services flow of productive services
  that capital delivers in production
• Measurement assumption capital services fixed
  proportion of productive capital stock

16
Productive capital stock
Efficiency and retirement parameter
Past investment
Results can be different from gross or net stock
Example Australia
17
Capital services and net capital stock, Australia
18
6. Interpretation of productivity measures

• Labour productivity how efficiently is labour
  used in production?
• MFP how efficiently are labour and capital
  combined used in production?
• MFP is residual --gt all those effects on output
  that are not investment in fixed capital
• Technical change, innovation
• Labour skills (unless measured in labour input)
• Economies of scale

19
Interpretation of productivity measures

• LP and MFP often pro-cyclical
• Measurement problem
• Capital (and labour) measures not adjusted for
  rates of utilisation
• Care with interpretation of year-to-year measures
• Puts a big question mark on quarterly
  productivity measures

20
7. OECD Productivity database

• Since March 2004, OECD productivity data
  available on the internet
• ProductivityStatistics Portal
• Labour productivity all OECD countries
• MFP 18 OECD countries
• Plan inclusion of some large non-member
  countries

21
8. Conclusions

• Productivity is a key indicator for analysis of
  economic growth --gt significant demand from
  policy makers
• Simple ratio but sometimes difficult to measure
• Sequence of development
• Top-down (start with total economy, total
  business sector)
• Labour productivity before MFP
• Annual measures rather than quarterly

22

Thank you for your attention!