Assessing the Reliability of
Labour Market Forecasts
Tony Meagher
Centre of Policy Studies, Monash University
International Symposium on Employment and Skills Forecasting,
University of Warwick, September 29, 2011
1
Plan of presentation
• Structure of the MONASH forecasting
system
• Reliability of the MONASH forecasts:
track record
• Reliability and the context of labour
market forecasting
2
Figure 4.
The MONASH
Forecasting
System for
Tight Labour
Markets
Macroeconomic
Forecasts
Structure of
Technical &
Taste
Changes (CoPS)
Expert
Opinion
MONASH
Simulations
2009-2010
to
2017-18
Labour Supply
Forecasts for
67 Skill Groups
Employment Forecasts for
214 Industries,
81 Occupations and
67 Skill Groups
Occupational Share
Effects (CoPS)
Labour Market Extension
I
Employment Forecasts for
358 Occupations
Qualification Share
Effects (CoPS)
Labour Market Extension II
Employment Forecasts for
7 Qualification Levels,
71 Qualification Fields
3
4
5
The MONASH track record
• Industries and occupations (Table 1)
• States and Territories (Table 2)
• Historical comparison (Table 4)
6
Comparative performance of MONASH
forecasts
• Trend extrapolation (Table 3)
• Alternative LFS sample (Table 1)
• Hodrick-Prescott filter (Table 3)
• Historical comparison revisited (Table 4)
7
School Education, Australia, Thousands
480
460
440
420
400
LFS data
380
HP filter
360
340
320
8
Machinery and Equipment Wholesaling, Australia,
Thousands
170
160
150
140
130
LFS data
HP filter
120
110
100
9
Computer Services, Australia, Thousands
190
170
150
130
110
LFS data
90
HP filter
70
50
30
10
Reliability in context I
Proposition 1. In choosing between different
forecasting methodologies, all alternatives should be
subject to the same criteria regarding reliability.
Proposition 2. If a workforce development strategy is to
equip the workforce with the skills it will need to meet
the future needs of industry, the strategy must identify
what the future needs of industry are going to be.
Proposition 3. In determining what skills are required to
meet the future needs of industry, structural linkages
are important.
11
Reliability in context II
Proposition 4. To the extent that the distribution of
training resources is left to market forces, labour market
forecasting has a valuable role to play in providing the
economic agents involved with information on which to
base their decisions.
Proposition 5. Whether or not a particular occupation
(or skill) constitutes a bottleneck to economic growth is
a property of the state of the economy and not a
property of the occupation.
12