The Challenge of Jobless Growth in Developing Countries: An Analysis with Cross-Country Data By Rizwanul Islam 28 February, 2010 Conference Room, BIDS.

The Challenge of Jobless Growth in Developing Countries: An Analysis with

Cross-Country Data

By

Rizwanul Islam

28 February, 2010

Conference Room, BIDS1

Content

• Background and purpose of the study

• Some conceptual clarifications

• Possibility of a trade-off between employment and productivity growth

• Employment and output growth: empirical evidence

• Decomposition of output growth into employment and productivity growth

• Constraints on employment growth

• Concluding observations2

The Background and Purpose of the Study

• Low and declining employment growth in relation to output

• Growth-poverty nexus:No invariant relationship between growth and

poverty reductionStrength of the nexus influenced by the

employment intensity of growth

• Hence the importance of addressing the issue of employment intensity of growth

3

Conceptual Clarifications

• What do we mean by jobless growth?Can output growth be jobless in a literal sense?Does the term “jobless growth” need to be

interpreted in a literal sense?

• Different possible combinations of output and employment growth

Low output low employment (stagantion)Low output high employment (growthless jobs)High output low employment (jobless growth)High output high employment (employment

intensive growth)

4

Combination of Output and Employment Growth

E

mpl

oym

ent G

row

th

Output Growth

I

High growth of employment and low

growth of output(growth-less jobs)

II

Low growth of output and employment

IV

High growth of output and of employment

(employment-intensive growth)

III

High growth of output with low growth of employment

(jobless growth)

5

Possibility of Trade-off between Employment and Productivity Growth

• Inverse relation between emp-elast and prody

• Hence the possibility of trade-off

• But both quantity of L input and prody can contribute to output growth

• Consider the following accounting identity∆Y = ∆ L + ∆ (Y/L)

• In a growing economy, both L and Y/L can grow

• The combination of L and Y/L would depend on a variety of factors

6

Employment and Output Growth: Empirical Analysis

• Focus on Manufacturing

• Why manufacturing?

• Data source(s): UNIDO industry data, and other data sources

• Periods: 1980s and 1990s

• Cross-country evidence on the relationship between employment and output growth

• Has there been a shift in the relationship between the two periods?

7

Relationship between employment and output growth (1980-89)

Annual Growth of Employment and Value Added: 1980-1989

Idn

MtiusBotw

KorBgd

SrlkMlsyPkst

KenThld

SAfr

Arg

IndMex

SgalPhil

Nig

Mlwi

y = 0.4458x - 0.0493

R2 = 0.2858

-10

-5

0

5

10

15

-15 -10 -5 0 5 10 15 20

Value added

Em

plo

ymen

t

Notes: Arg: Argentina; Bgd: Bangladesh; Botw: Botswana; Ind: India; Idn: Indonesia; Ken: Kenya; Kor: Korea; Mlwi: Malawi; Mlsy: Malaysia; Mtius: Mauritius; Mex: Mexico; Nig: Nigeria; Pkst: Pakistan; Phil: Philippines; Sgal: Senegal; Safr: South Africa; Srlk: Sri Lanka; Thld: Thailand.Source: Author's calculations based on Unido, Indstat 3, 2005. 8

Relationship between Employment and Output Growth (1990-2002)

Annual Growth of Employment and Value Added: 1990-2002

PhilKor

MtiusPkstSafr

ArgInd

Sgal

BotwIdnGhn

Mlwi

Ken

MlsySrlk

MexBolv

BgdThld

Nig

y = 0.1561x + 2.5666

R2 = 0.0212

-5

0

5

10

15

20

-5 0 5 10 15 20Value added

Em

ploy

men

t

Notes: Arg: Argentina; Bgd: Bangladesh; Bolv: Bolivia; Botw: Botswana; Ghn: Ghana; Ind: India; Idn: Indonesia; Ken: Kenya; Kor: Korea; Mlwi: Malawi; Mlsy: Malaysia; Mtius: Mauritius; Mex: Mexico; Nig: Nigeria; Pkst: Pakistan; Phil: Philippines; Sgal: Senegal; Safr: South Africa; Srlk: Sri Lanka; Thld: Thailand.Source: Author's calculations based on Unido, Indstat 3, 2005.

9

Employment and Output Growth in Selected Countries of Asia

Growth of Manufacturing Output and Employment in Selected Countries of Asia

Indonesia (2000-03)

China (1990-95)India (1993-99)Thailand (2001-04)

China (1985-90)Sri Lanka (1990s)

India (1983-93)

India (2000-2005)Sri Lanka (1980s) China (1995-99)

Malaysia (1995-2000)Thailand (1980-89)

Thailand (1990-96)

Bangladesh (1991-2002)

Indonesia (1975-96)Cambodia (1994-2004)

Malaysia (1990-95) Malaysia (1985-90)

y = 0.4607x + 0.207R2 = 0.5722

-2

0

2

4

6

8

10

12

14

0.0 5.0 10.0 15.0 20.0 25.0 30.0

Output

Empl

oym

ent

Source: Except Bangladesh, the figures from other countries have been obtained from in-depth country level studies that are referred to in Khan (2007). The figures for Bangladesh are from Ahmed, et al. (2009).

10

Relationship between Employment and Output Growth: Overview by Region

• Asia:High output and employment growth: Bangladesh,

Indonesia (1980s and first half of 1990s), Malaysia, Thailand

High output growth with low employment growth: China, India, Indonesia (after economic crisis)

Decline in the employment intensity of growth: China, India

Bangladesh: Contradictory evidence:UNIDO data show improvementEarlier studies (e.g., ADB, 2005, Rahman and Islam,

2006)show decline in employment intensity 11

Employment and Output Growth: Overview by Region (contd.)

• Sub-Saharan Africa:Decline in employment intensity during the 1990s

compared to the 1980s: Botswana, Malawi and Mauritius

Improvement in employment intensity of growth: Nigeria and Senegal

Negative employment growth with positive output growth: South Africa (1990-02)

12

Employment and Output Growth: Overview by Region (contd.)

• Middle East and North AfricaLarge variation: emp-elast. 0.14 in Egypt and 1.27 in

JordanHigh employment elasticity (0.5-0.8): Morocco, Syria,

Tunisia and YemenLow output growth (less than 5%) is a more serious

problem: (Algeria, Egypt, Iran, Morocco, and Tunisia)

• Latin AmericaArgentina and Mexico: very little change in employment

intensity of growthBrazil: decline in employment intensity during the 1990s

compared to the 1980s.13

Decomposition of Output Growth into Employment and Productivity Growth

• Periods: Manufg : 1980-89 and 1990-2002 Overall GDP: 1980-90, 1990-96, 1996-2002

• Balanced contribution of employment and productivity growth: Korea, Malaysia, Taiwan, Thailand

• In South Asia, the contribution of labour productivity growth is higher than expected

• The contribution of Labour productivity increased in Bangladesh, China, India, Indonesia, Malaysia

• The pattern of growth in South Asia and China less conducive to employment growth compared to East and South East Asia

14

Constraints on Employment Growth

• The notion of “binding constraints” used in growth diagnostic literature (Hausman, Rodrik and Velasco)Marginal welfare benefit of reducing distortion

• Possible constraints on employment growthThe neoliberal approach and the blame on labour

market interventionsChoice of technology and capital deepeningGoing beyond technology: start from the pattern of

growth

15

Constraints on Employment Growth (contd.)

• There is no conclusive evidence on the adverse effects of LM interventions on employmentBean (1994, JEL): Evidence do not show that

generous unemployment benefits in Europe was the cause of persistent unemployment

Nickel (1997, JEP): Unemployment benefits do not have an adverse effect on unemployment rates

Forteza and Rama (2002, ADB report 2005) covering 119 countries: minimum wages and mandated benefits do not hinder economic growth

16

Constraints on Employment Growth (contd.)

• Kapsos (2005, ILO WP): Rigidities in the labour market do not have a negative effect on employment elasticity The coefficient of the World Bank’s employment

rigidity index is not statistically significant and the sign is opposite (cross section, 100 countries) Rigidity of employment index is the average of

three indices: difficulty of hiring difficulty of firing rigidity of hours

17

Pattern of Growth and Employment

• Sector composition of output: Manufacturing and others Growth of manufacturing in relation to GDP growth has been

much higher in countries of ESEA compared to those in South Asia

• Sector composition of output: within manufacturing In Malaysia and Korea, share of labour intensive industries

increased up to 1990 In Thailand, this happened till 2002 India witnessed a decline in the share of labour-intensive

industries

18

Pattern of Growth and Employment (contd.)

• Bangladesh:The share of both top five labour intensive and top

five capital intensive industries increasedThe share of ready made garments increased very

rapidlyThe share of a major labour intensive industry, viz.,

leather and leather products declined

• Sri Lanka:Similar to Bangladesh

19

Pattern of Growth and Employment (contd.)

• Pattern of demand: external Theory of comparative and the impact of trade openness

India: Trade openness has not resulted in an increase in the share of labour intensive sectors in exports

Pakistan: Increase in the share of capital intensive exports alongside high growth labour intensive exports

• Pattern of demand: domestic India: expenditure elasticity of demand for labour intensive

products higher for lower and middle income growps Difference in average expenditure on consumer durables more

marked than for basic items

20

Pattern of Growth and Employment (contd.)

• Pattern of demand: domesticBangladesh: Income elasticity of demand for labour

intensive goods lower for richer peopleSome labour intensive goods (e.g., gur, firewood) are

“inferior goods” for the richer people(See Table in the following slide)

• Implication of the domestic pattern of demand: Importance of the level of income of the poor and the distribution of income

21

Income Elasticity of Demand for Selected Consumer Goods in Bangladesh (1005-06)

Items Income elasticity of demand

Top 10% of the households

All households

Gur -0.001 0.164

Firewood -0.031 0.179

Lungi 0.100 0.231

Shirt and pant 0.353 0.616

Mill made cloth 0.149 0.306

Handloom cloth 1.453 0.338

Leather shoes 0.328 0.516

Kitchen items 0.502 0.438

Regrigerator, pressure cooker, etc.

1.169 0.566

Furtniture 0.623 0.743Note: Estimated from Household Income and Expenditure Survey data of 2005-06. 22

Concluding Observations

• The term jobless growth should not be interpreted literally

• Employment intensive growth implies high growth of both output and employment

• The relationship between employment and output growth is not invariant

• Empirical evidence points to a decline in the employment intensity of growth in many developing countries

• Growth in ESEA has been more employment intensive than in South Asia

• In some countries (e.g., China and India), high output growth associated with low and declining employment intensity

23

Concluding Observations (contd.)

• In a dynamic economy, employment and labour productivity can grow together Countries of ESEA has done better than those of South Asia

• Pattern of growth (sector composition) is important for employment intensive growth

In countries of ESEA, sector composition has been more conducive to employment

Trade openness does not necessarily lead to an employment intensive pattern of growth

Income distribution and the income of the poor are important from the point of view of growth of labour intensive sectors

24

Concluding Observations (contd.)

Policy Importance of overall policies shaping the pattern of

income distributionPossible distortions in incentive structure

Capital made cheaper (India) Policy environment (including stimulus measures)

encouraging the consumption of capital-intensive items (China, Indonesia)

Absence of any measure to encourage the use of labour

Identification of constraints operating on the supply side for the labour intensive sectors

Possibility of positive support

25