Structural Change: Implications of Policy and other Barriers

1
Structural Change Implications of Policy and
other Barriers

• Christopher A Pissarides
• Centre for Economic Performance
• London School of Economics
• Presentation at Oxonia
• 10 May 2005

2
Contents

• This presentation is based on a joint project
  with Rachel Ngai of the London School of
  Economics, financed by the ESRC
• Motivation for the study
• Structural change and economic growth
• Barriers to structural change
• Policy implications

3
Motivation

• Growth takes place at uneven rates across
  industrial sectors
• This usually gives rise to structural change
• Defined as the movement of labour and capital
  across industrial sectors

4

• What are the implications of uneven growth for
  structural change when there are no frictions?
• What are the implications of mobility barriers
  and frictions?
• Is there evidence that policy or other barriers
  matter?

5
Our Modelling Approach

• Economy consists of many distinct sectors
• Some produce final consumption goods (e.g.
  clothes and food)
• Some produce both consumption goods and capital
  or intermediate goods (e.g. manufacturing)
• And some production takes place at home (e.g.,
  cleaning and ironing)

6

• TFP growth is unbalanced new technology is not
  uniformly spread across industrial sectors
• We claim that unbalanced TFP growth is the cause
  of structural change
• When does structural change take place and how?

7
Nature of structural change

• Structural change requires non-unitary income or
  price elasticities
• Otherwise differences in TFP growth rates are
  absorbed by prices
• Our model implies unit income elasticities, so
  structural change is due to non-unitary price
  elasticities

8

• Formally, we have
• constant elasticity of substitution utility
  functions defined over consumption goods (and
  leisure)
• Cobb-Douglas production functions in all sectors
• one aggregate capital good allocated to all
  sectors
• fixed labour force allocated to all sectors
• or, fixed total annual hours allocated to market
  work, home work and leisure

9
Implications

• If consumption goods are not close substitutes
  (elasticity of substitution less than 1)
• labour moves to sectors with low TFP growth
• prices of goods with low TFP growth rise
• real consumption shares are approximately
  constant
• There is a limiting state with only two sectors,
  the slowest growing consumption sector and the
  capital-producing sector

10

• These are more of less the implications of
  Baumols 1967 classic Unbalanced Growth model
• labour moves to stagnant sectors
• stagnant sectors suffer from Baumols cost
  disease
• economys growth path is on a declining trend

11

• But crucially, we show that without mobility
  barriers economy is on a steady-state growth path
• Growth rate equal to the rate of growth of the
  capital-producing sector
• If there are mobility barriers structural change
  is slower and aggregate growth rate not strictly
  constant

12
Parameter restrictions

• Elasticity of substitution across broad sectors,
  e.g., two-digit industries, less than 1 (their
  products are poor substitutes), e.g.,
• Food, clothes, TV sets
• But elasticity of substitution within commodity
  groups is 1 or bigger, e.g.,
• whether you use electronic means to store
  information or paper means
• whether you eat at home or go to restaurant

13
Implications

• At the two-digit level or broader
• labour share of slow-growing consumption sectors
  is expanding
• labour share of fast-growing consumption sectors
  is contracting
• labour share of capital goods production
  converging to investment share of output
• real consumption shares approximately constant

14

• Within narrower groups,
• Labour share of fast TFP-growth sectors
  expanding, e.g., ICT, types of cloth sold
• Home production time declining if market
  technology grows faster than home technology

15
Employment flows during structural change
Labour flow
16
International trade

• Tradable goods have good substitutes across
  nations
• Hence, fast TFP-growth sectors producing tradable
  goods may retain or attract labour share
• Provided they grow faster than international
  competitors

17
Barriers to economic activity

• Two types, mobility barriers and
  taxation-regulation of activities
• Effect of mobility barriers temporary but because
  structural change is ongoing, they last for a
  very long time
• Effects of regulation-taxation may persist
  indefinitely

18
Taxation-regulation

• Taxation-regulation of market activity drives
  work hours to the home
• It affects mainly market sectors that are close
  substitutes to home production (e.g., services)
• May explain some of the gap across nations in
  employment rates, especially in services

19
Taxation-regulation and TFP

• Think of taxation-regulation as having opposite
  effect of TFP
• Prescott and Rogerson ask, can it explain the
  difference between US and EU employment rates?
• Main differences should be in womens employment
  and in services

20

• Reverse engineering exercises difference in
  employment rates can be explained by a 40
  effective shortfall in European TFP.
• Given the much smaller measured shortfall, rest
  is attributed to taxation-regulation
• But this concerns level of TFP and substitutions
  between home/market
• Does not need dynamics

21

• Except that over time, if TFP growth in market is
  higher than in home, effect of taxation-regulation
  should be getting smaller as home hours fall
• So eventually negative implications of taxation
  should be diminishing

22
Mobility barriers

• For capital the cost of closing down businesses
  in contracting sectors and setting up businesses
  in new sectors
• For labour policies that inhibit mobility, e.g.,
  social security, housing, etc.
• Education, training, lifelong learning
  facilities, adaptability

23
Implications

• Barriers to closing down jobs and businesses
  likely to slow down structural change
• But are likely to be less important than barriers
  to expansion in new sectors, because the normal
  turnover of labour eventually allows decline

24

• Therefore mobility barriers are likely to have
  bigger impact on expanding sectors than on
  contracting sectors
• There should be more unemployment in growth
  equilibrium
• Less job creation in services and other expanding
  sectors
• For example, less fast uptake of ICT production
• More home production

25
Data

• In the process of compiling summary institutional
  data on barriers
• I give the ranking of countries on the basis of
  barriers to entrepreneurship, very similar to
  ranking on the basis of employment protection
  legislation

26
Ranking of countriesAdmin burden on start-ups
1 Denmark 11 Finland
2 US 12 Portugal
3 UK 13 Japan
4 New Zealand 14 Greece
5 Ireland 15 Austria
6 Australia 16 Germany
7 Sweden 17 Belgium
8 Canada 18 Spain
9 Netherlands 19 France
10 Norway 20 Italy
27
More on data

• Data on 2-digit sectors but for the next set of
  graphs aggregated into five sectors
• Agriculture and mining
• Manufacturing
• Other production industries
• Business services
• Personal services (including government)

28
Model implications I

• TFP growth in agriculture and manufacturing
  higher than in services
• Employment share of agriculture falls towards 0
• Share of manufacturing falls towards investment
  rate without trade
• With trade it may rise or remain above investment
  rate
• Share of services rises towards consumption rate

29
(No Transcript)
30
(No Transcript)
31
(No Transcript)
32
(No Transcript)
33
(No Transcript)
34
(No Transcript)
35
Model implications II

• TFP growth in the market higher than in the home
• Female employment should be rising

36
(No Transcript)
37
Model implications III

• More mobility barriers
• Share of services lower
• Share of manufacturing out of working age
  population falling but share of services not
  rising fast enough to compensate
• Unemployment rate higher
• Employment-to-population rate lower

38
Manufacturing employment share of WAP
39
Service share of WAP
40
US
UK
FR
GE
IT
41
Implications of last three charts

• Manufacturing share in working age population
  converging
• Service share not converging
• Therefore, net transfer of working time from
  manufacturing share to out of the labour force
• Barriers affect job creation in services
  adversely, which translates to lower overall
  employment

42
Services share of WAP vs. barriers, 17 countries
43
Manufacturing share of WAP vs. barriers
44
Service share vs rank of barriers, 17 countries
45
Gender employment gap vs. barriers rank
46
Is service share the only problem for women?

• Very preliminary US service expansion seems to
  be creating more jobs for women than expansion of
  service share in Europe
• Justified by our claim that barriers, in addition
  to slowing down convergence, drive production to
  the home

47
Service employment share and womens employment
1970-2000
48
Conclusions I

• Normal process of economic development requires
  structural change because of unbalanced TFP
  growth
• Delaying the process (sometimes called
  deindustrialisation) harmful
• May temporarily hold average TFP growth up but
  eventually structural change will dominate

49
Conclusions II

• Low female employment in countries like Italy and
  Spain seem to be associated with high barriers to
  change
• Low service employment share also symptom of
  barriers
• Barriers have bigger impact on expanding service
  share than on contracting manufacturing share