Transcript Using National Accounts Data for Productivity Analysis

Using National Accounts Data for
Productivity Analysis
François Lequiller
OECD
1
Objectives of the paper

Promote the use of national accounts data on labour input for
productivity analysis

The paper
– makes several recommendations regarding the transmission of
data to OECD/Eurostat
– Proposes a model questionnaire on the bridge table between labor
force statistics and national accounts data on employment
2
Productivity comparisons in the OECD

To compare productivity growth is easier than to
compare productivity levels (PPPs…)

However, international users want to compare levels

Example: EU countries have set benchmark to the
US in terms of levels (Lisbon objectives)

Canada and US productivities are compared in levels

OECD disseminates data on productivity levels
3
Australia
Canada
France
Germany
Italy
Japan
Sweden
United Kingdom
United States
Euro-area
GDP per
head of
population
(as % of US)
GDP per
hour worked
(as % of US)
76
85
77
75
75
74
74
74
100
73
77
85
113
93
94
71
86
79
100
92
4
Decomposition of GDP per capita

GDP/Population = (GDP/total hours worked) * (total hours
worked/persons employed) * (persons employed/labour force) *
(labour force/working age population) * (working age
population/population)

Definition of labour productivity = GDP/total hours worked, in
preference to GDP/persons employed

Reason: total hours worked vary a lot between OECD countries
5
Limitations to GDP/total hours worked

Does not take into account differences in the
composition of the work force

Difficulty to measure the non market sector

UK Atkinson report

Productivity of business sector better measured

But current limitation of the international data
available on business sector
6
Use of national accounts data

The numerator of labour productivity is GDP

It is natural that the denominator, the measure of labour input, is
consistent with numerator
Recommendation 1: national accounts data should be used to
compile the denominator of labour productivity
Recommendation 2: standard national accounts tables should
allow for the calculation of the productivity of the non farm
market sector
Market = Total – NPISH – General Government – Owner
occupied housing



7
Conditions to use national accounts data

First condition: total hours data in the framework of national
accounts should be compiled by countries and transmitted to the
OECD

Second condition: they should be comprehensive and
consistent with GDP (coverage of self-employed, underground
economy, domestic concept)

Third condition: they should be comparable between countries
8
Availability of national accounts data

See Table 2, page 6

Most OECD countries transmit data on employment

However, some transmit numbers in terms of jobs, while the
majority transmits data in terms of persons

Only very few countries transmit data on total hours

Some transmit data on total hours for employees but not for selfemployed
9
Improve international availability of data on total
hours

Recommendation 3: OECD countries should report
systematically total hours for employees and self-employed

The concept of full-time equivalent is being abandonned

Recommendation 4: SNA should be clarified to avoid excluding
the concept of « persons ». ESA could be used as an input to
precise what is in or out of labor input.
Paris Group (labour force statistics experts) are discussing of a
new ILO resolution on working time measurement. Could lead to
some precisions in the SNA paragraphs that refer to the old
resolution.
Recommendation 5: OECD countries should report data in
terms of persons as well as in terms of jobs.


10
Comparability with labour force statistics

See table 3, page 9 (employment)

See table 4, page 10 (hours worked)

Significant differences with original source
Some differences can be explained
Example: US: jobs (NIPA, from CES) versus persons (CPS)
Example: Italy: partially underground economy
More information is needed to better understand and to check
that national accounts adjustments are comparable




11
From a pilot test to a questionnaire




Objective: better understand the bridge table
between original labour force sources and national
accounts data
France and Canada have accepted to describe this
bridge table
Based on this experience, Eurostat and OECD
propose a questionnaire quantifying the adjustments
made
The two pilot countries are illustrative:
– France: uses mainly Census, administrative data and
business surveys and introduces corresponding adjustments
– Canada: uses mainly LFS, and introduces corresponding
adjustments
12
OECD/Eurostat questionnaire

Objective: explain and quantify the differences
between the main labor force statistics and the NA
estimates

Start from the labour force statistics
– For persons/jobs
– For hours worked

Quantify each adjustment
13
OECD/Eurostat questionnaire

Quantify adjustments for employment
– From stocks to flows (averaging)
– From persons/jobs to persons/jobs
– Adjustments for coverage
• Military
• Other collective households
• Territories
• Residents working outside (-)
• Non residents working inside (+)
– Unobserved economy
– Other
14
OECD/Eurostat questionnaire (ctd)

Quantify the adjustments introduced for hours worked
–
–
–
–
–
For annual leaves and holidays
For sickness leaves
For strikes and temporary lay-offs
For paid but unreported overtime
For unpaid overtime
15
Conclusion
The objective is to be able in end 2005:
– To use at OECD national accounts based numbers for the
denominator of the headline labour productivity level number
– To be able to explain the source of the data, their difference
with labour force statistics, and their international
comparability
Comments on the recommendations and on the
questionnaire are welcome.
16