S KILLS M ISMATCH AND I NFORMAL S ECTOR P ARTICIPATION AMONG E DUCATED I MMIGRANTS : E VIDENCE FROM S OUTH A FRICA Alexandra Doyle Amos Peters Asha Sundaram.

SKILLS MISMATCH AND INFORMAL SECTOR PARTICIPATION AMONG

EDUCATED IMMIGRANTS:

EVIDENCE FROM SOUTH AFRICA

Alexandra Doyle

Amos Peters

Asha Sundaram

Presented at

HSRC LMIP Seminar Series

October 30, 2014

2

AIM We look at employment outcomes for immigrants relative to

natives in the SA labour market.

Particularly, we show that:

conditional on education, immigrants from different countries have different likelihoods of being employed

in a skilled job in the informal sector

Argue that this is evidence for ‘brain waste’.

We study correlations between outcomes and origin-country characteristics

3

MOTIVATION

Immigrants can contribute to the host country labour market by bringing:

Diversity Skills

Poor absorption of immigrants into the labour market can lead to:

‘brain waste’: relevant for developing host countries expanding informal sector/unemployment issues like increasing crime, related to the above

4

LITERATURE

Large literature on immigrant performance and assimilation in host country labour markets, focusing on employment, wages.

Borjas, 1994; Borjas, 2003; Ottaviano and Peri, 2012

It is important to look at quality of jobs.

Mattoo, Neagu & Ozden, 2008; Bourgeault et al., 2010; Carr, Inkson & Thorn, 2005

Literature has focused on developed host countries.

5

CONTRIBUTION

We focus on ‘brain waste’ in SA - a developing country. This is important as:

Greater market imperfections make finding the right job more difficult in countries like SA

High unemployment Existence of an informal sector Rigid labour market Discrimination

Shortage of skills means immigrant opportunities differ, skills utilisation more important.

6

IMMIGRANTS AND LABOUR MARKET IN SAo Immigration policy was de-racialized in 1994.

o SA is a regional economic power, attracting skilled immigrants from other African countries.

o SA has a diverse immigrant pool from OECD and other African countries.

o However, its labour market is rigid, and high unemployment persists.

o Historical legacy means that SA achieving employment equity goals remains a huge concern.

7

IMMIGRANTS AND LABOUR MARKET IN SAo SA ranks high in anti-immigrant sentiment.

o Facchini, Mayda and Mendola, 2011

o Foreign immigrants from different countries ‘perceived’ differently.o Southern African Migration Project

Favourable

(%)

Whites Blacks Coloured

s

Asians/Indians Total

Nigerians 11 8 4 9 8

Angolans 14 9 5 7 9

Batswana 29 40 14 23 35

DRC 15 10 5 6 10

Ghanaians 16 12 4 9 11

Basotho 27 46 17 23 39

Mozambicans 13 15 9 11 14

Somalis 9 10 5 17 10

Swazi 24 44 18 32 38

Zimbabweans 12 13 9 11 12

8

IMMIGRANTS AND LABOUR MARKET IN SA

o There is recognition in the literature that skilled African immigrants potential solution to skills shortage.

o There is also evidence that immigrants contribute to local economy.

o However, immigration policy remains partial to immigrants from advanced economies.

o This might hamper utilisation of African immigrant skills.o Rasool, Botha and Bisschoff, 2012; Kalitanyi and Visser, 2010; Mattes,

Crush and Richmond, 2000; Peberdy, 2001

9

PREVIEW OF RESULTS

Brain waste? Yes!

Substantial variation (by country of origin) among educated immigrants in likelihood of finding a skilled job

Immigrants from most African countries have lower likelihood of obtaining a skilled job relative to natives

Educated migrants from W. Africa, Kenya, Ethiopia, Somalia, Sudan and Eritrea have a high probability of obtaining an informal sector job

Positive Negative

Obtaining a skilled job

GDP per capita, schooling quality

Presence of conflict

Obtaining an informal sector job

Presence of conflict, Distance to SA

English official language, GDP per capita

10

FRAMEWORK

Logit model to estimate probability of being in high-skilled (versus low-skilled) job:

Xi is a vector of individual characteristics:

highest education level Age, age squared duration in province, urban/rural, marital status, origin-country group

11

FRAMEWORK We obtain predicted probabilities of obtaining a high-skilled

job for each origin-country group

Probabilities are relative to the benchmark: Native internal migrants Better comparison group, since they are likely to be

positively selected on unobserved factors

For a subset of non-OECD country groups: We estimate the probability of being employed in a

skilled, unskilled formal and unskilled informal job

12

FRAMEWORK

Multinomial logit model for outcome j:

Yi : Probability of obtaining a formal skilled, formal unskilled or informal job

Xi is a vector of individual characteristics

13

DATA

We use South African census data for 2001.

We look at 30 origin-country groups and native internal migrants.

Occupations classified into skilled (professional, skilled, semi-skilled), unskilled.

We exclude farming activities and the unemployed.

14

SAMPLE

Males, between ages 25 and 65

Recent migrants: arrived in current province between 1996 and 2001

For immigrants:

Restrict to those who arrived at an age where education was likely to have been obtained abroad

15

IMMIGRANTS IN SA

Continents of origin (%)

AfricaNorth AmericaEuropeAsiaAustralasia

16

DATA DESCRIPTIONImmigrant

s  Internal Migrants

Sample Size 3 919   45 276

Distribution by Educational

Attainment 

No Schooling 13.2%   6.9%

Primary School 17.6%   18.7%

High School (matric equivalent) 44.6%   57.3%

Undergraduate Degree or Diploma 16.0%   13.7%

Masters or Doctorate Degree 8.7%   3.3%

Distribution by Occupation Type      

Formal Sector: 77.4%   88.3%

- Unskilled Worker 50.1%   59.3%

- Skilled Worker 27.3%   29.0%

Informal Sector: 22.6%   11.7%

- Unskilled Worker 19.8%   10.1%

- Skilled Worker 2.8%   1.6%

17

RESULTS: PROBABILITY OF OBTAINING SKILLED EMPLOYMENT

Country/Region

Undergraduat

e Postgraduate Country/Region

Undergraduat

e Postgraduate

RSA 73.6 89.6

Namibia & Botswana 80.5 92.7 China 78.5 91.9

Lesotho 47.3 73.5 East Asia 84.9 94.6

Zimbabwe 73.4 89.5 South East Asia 48.6 74.5

Mozambique 55.3 79.3 Bangladesh, Nepal, SL 65.8 85.6

Swaziland 63.2 84.1 India 73.0 89.3

Angola 68.4 87.0 Pakistan 54.8 78.9

Malawi & Zambia 68.0 86.7 North America 73.9 89.7

Congo & Gabon 57.2 80.5 Australasia 94.4 98.1

DRC & Cameroon 53.1 77.8 UK & Ireland 89.8 96.5

Tanzania 76.8 91.1 Western Europe 92.7 97.5

Kenya 66.9 86.2 Germany & Austria 79.0 92.1

Burundi, Rwanda & Uganda 72.1 88.9 Eastern Europe 74.6 90.1

East Africa 62.8 83.9 Mediterranean Europe 67.6 86.6

Nigeria 65.7 85.5 Scandinavia 61.8 83.3

West Africa 42.9 69.9

North Africa 93.5 97.8

18

RESULTS: PROBABILITY OF OBTAINING UNSKILLED INFORMAL EMPLOYMENT

Undergraduate Country/region Skilled Formal Informal

RSA 73.6 25.0 1.4

Lesotho 47.6 49.2 3.2

Namibia & Botswana 80.3 18.7 1.0

Zimbabwe 76.5 19.5 4.0

Mozambique 56.6 40.0 3.4

Swaziland 62.8 35.6 1.6

Angola 71.3 24.0 4.7

DRC & Cameroon 55.7 36.4 7.8

Congo & Gabon 58.5 36.9 4.6

Malawi & Zambia 69.5 26.7 3.8

Tanzania 77.8 20.5 1.8

Burundi, Rwanda & Uganda 73.5 21.9 4.6

East Africa 68.2 21.7 10.2

West Africa 46.3 34.7 19.0

Kenya 69.4 19.0 11.7

Nigeria 69.6 20.5 9.9

Bangladesh, Nepal & Sri Lanka 67.5 29.0 3.4

India 74.8 19.1 6.0

Pakistan 56.5 36.6 6.9

19

RESULTS: PROBABILITY OF OBTAINING UNSKILLED INFORMAL EMPLOYMENT

Postgraduate

Country/region Skilled Formal Informal

RSA 89.7 9.9 0.4

Lesotho 74.0 24.8 1.2

Namibia & Botswana 92.7 7.0 0.3

Zimbabwe 91.3 7.5 1.2

Mozambique 80.4 18.4 1.2

Swaziland 84.1 15.4 0.5

Angola 88.9 9.7 1.4

DRC & Cameroon 80.3 17.0 2.7

Congo & Gabon 81.7 16.7 1.6

Malawi & Zambia 87.9 10.9 1.2

Tanzania 91.7 7.8 0.5

Burundi, Rwanda & Uganda 90.0 8.7 1.4

East Africa 87.8 9.0 3.2

West Africa 74.5 18.0 7.4

Kenya 88.5 7.8 3.6

Nigeria 88.5 8.4 3.1

Bangladesh, Nepal & Sri Lanka 86.8 12.1 1.1

India 90.7 7.5 1.8

Pakistan 80.7 16.9 2.4

20

FIGURE: GRADIENTS

Undergraduate Postgraduate 0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Skilled Employment-Education Gradient

RSALesothoMozambiqueNamibia & BotswanaWest AfricaKenyaNigeriaChinaUK & Ireland

Education Level

Prob

abil

ity

21

FIGURE: GRADIENTS

Undergraduate Postgraduate0%

2%

4%

6%

8%

10%

12%

14%

16%

18%

20%

Informal Employment-Education Gradient

RSAWest AfricaKenyaNigeriaIndiaDRC & CameroonNamibia & Botswana

Education Level

Prob

abil

ity

22

CORRELATIONS: PROBABILITY OF SKILLED FORMAL EMPLOYMENT WITH ORIGIN COUNTRY CHARACTERISTICS

Undergraduate Degree Postgraduate Degree

Log of Distance to SA 0.28 0.26

Military Conflict -0.50 -0.53

Asylum/Refugee applications -0.24 -0.21

English 0.04 0.04

Log of GDP per capita 0.53 0.51

Pupil Teacher Ratio -0.38 -0.35

Source: WDI, CIA Factbook, www.prio.no, La Porta and Schleifer (2008), United Nations, CEPII

23

CORRELATIONS: PROBABILITY OF INFORMAL EMPLOYMENT WITH ORIGIN COUNTRY CHARACTERISTICS

Undergraduate Degree Postgraduate Degree

Informality 0.16 0.15

Log of Distance to SA 0.57 0.53

Military Conflict 0.51 0.49

Asylum/Refugee applications 0.42 0.33

English -0.57 -0.53

Log of GDP per capita -0.39 -0.34

Pupil Teacher Ratio 0.25 0.16

Source: WDI, CIA Factbook, www.prio.no, La Porta and Schleifer (2008), United Nations, CEPII

24

FIGURE: EMPLOYMENT GRADIENTS

No Schooling High School Undergraduate Postgraduate0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Education-Employment Gradients

South AfricaDRCMalawiCameroonNigeriaIndiaUK, Ireland

Pro

babilit

y o

f Em

plo

ym

ent

25

SUMMARY OF RESULTS There is substantial variation (by country of origin) among educated

immigrants in the likelihood of finding a skilled job.

Immigrants from most African countries have a lower likelihood of obtaining a skilled job relative to natives.

Educated migrants from W. Africa, Kenya, Ethiopia, Somalia, Sudan and Eritrea have a high probability of obtaining an informal sector job.

We see that the variation in probabilities is lower at higher levels of education.

Source country characteristics are correlated with immigrant performance.

Positive Negative

Obtaining a skilled job

GDP per capita, schooling quality

Presence of conflict

Obtaining an informal sector job

Presence of conflict, Distance to SA

English official language, GDP per capita

26

IMPLICATIONS

Policy Implications:

Streamline immigration policy

Aid immigrant assimilation

Access to better information for employers

Easier accreditation

27

THANK YOU

28

Variable Definition Source

(log) Distance to South

Africa

Kilometres between capital city of immigrant source

country and South AfricaCEPII’s Distance Measures

Informality % of the active labour force self-employed

La Porta and Schleifer (2008), accessed at

http://faculty.tuck.dartmouth.edu/rafael-laporta/res

earch-publications

Refugee ApplicationsRatio of refugees and asylum seekers to total immigrant

stock in 2000United Nations Commission for Refugees

Military conflictA dummy variable which takes on the value 1 if there was

military conflict in the home country during 1996-2001

Variable constructed using www.prio.no, version

2.1 of the “Armed Conflict” database initiated by

Gleditsch, Wallensteen, Eriksson, Sollenberg and

Strand (2002)

EnglishEnglish as an official language--dummy variable with

value 1 if English is the official spoken languageCIA - The World Factbook, 2014.

Pupil-Teacher ratio Number of pupils to teachers in an average class, 2001 World Development Indicators.

(log) GDP per capita per capita GDP adjusted for PPP, 2001 World Development Indicators.

29

COUNTRY GROUPSLesotho

Namibia Botswana

Zimbabwe

Mozambique

Swaziland

Angola

Democratic Republic of The Congo (Zaire) Cameroon

Congo Gabon

Malawi Zambia

Tanzania

Algeria Libya Egypt Iran Israel Jordan Lebanon Turkey Morocco

Burundi Rwanda Uganda

Eritrea Ehiopia Somalia Sudan

Ghana Benin Cote d'ivoire Sierra Leone Liberia Senegal

Kenya

Nigeria

US Canada

China HK

Bangladesh Nepal Sri Lanka

India

Japan South Korea North Korea Taiwan

Malaysia Philippines Singapore Indonesia

Pakistan

UK Ireland

Bulgaria Croatia Russia Poland Slovkia Macedonia Yugoslavia Ukraine

Denmark Finland Netherlands Norway Sweden

France Belgium Switzerland

Germany Austria

Portugal Italy Greece Spain Cyprus

Australia New Zealand