women and men in the informal economy : a statistical picture · picture, using the available data,...

WOMEN AND MENIN THE INFORMAL ECONOMY :

A statistical picture

INTERNATIONAL LABOUR OFFICE GENEVA

EMPLOYMENT SECTOR

Copyright © International Labour Organization 2002First published 2002

Publications of the International Labour Office enjoy copyright under Protocol 2 of the Universal Copyright Conve n t i o n .Nevertheless, short excerpts from them may be reproduced without authorization, on condition that the source is indicated. Forrights of reproduction or translation, application should be made to the Publications Bureau (Rights and Permissions),International Labour Office, CH-1211 Geneva 22, Switzerland. The International Labour Office welcomes such applications.Libraries, institutions and other users registered in the United Kingdom with the Copyright Licensing Agency, 90 TottenhamCourt Road, London W1T 4LP [Fax: (+44) (0)20 7631 5500; email: [email protected]], in the United States with the CopyrightClearance Center, 222 Rosewood Drive, Danvers, MA 01923 [Fa x: (+1) (978) 750 4470; email : [email protected]] or inother countries with associated Reproduction Rights Organizations, may make photocopies in accordance with the licencesissued to them for this purpose.

ISBN 92-2-113103-3

The designations employed in ILO publications, which are in conformity with United Nations practice, and the presentation ofmaterial therein do not imply the expression of any opinion whatsoever on the part of the International Labour Office concerningthe legal status of any country, area or territory or of its authorities, or concerning the delimitation of its frontiers.The responsibility for opinions expressed in signed articles, studies and other contributions rests solely with their authors, andpublication does not constitute an endorsement by the International Labour Office of the opinions expressed in them. Reference to names of firms and commercial products and processes does not imply their endorsement by the InternationalLabour Office, and any failure to mention a particular firm, commercial product or process is not a sign of disapproval.

ILO publications can be obtained through major booksellers or ILO local offices in many countries, or direct from ILOPublications, International Labour Office, CH-1211 Geneva 22, Switzerland. Catalogues or lists of new publications are ava i l a b l efree of charge from the above address.

Printed in France BRI/SAD

The purpose of this report is to provide a statisticalpicture, using the available data, of the informaleconomy worldwide. The statistics and indicatorspresented in this book are designed to give a wideaudience of users a better understanding of thesize, components and characteristics of theinformal economy. Such information is essentialin formulating policies and programmes at nationaland international levels to promote decent condi-tions of work as well as contribute to poverty erad-ication.

The statistics in the report were analysed andpresented according to a new conceptual frame-work for the informal economy proposed in thereport on Decent Work and the Informal Economyfor general discussion at the 90th Session of theInternational Labour Conference, Geneva June2002. The resulting analysis shows the feasibilityand usefulness of the framework as a basis for thed evelopment of statistics and for policy analysis.N o t a b l y, in the three countries featured as casestudies, the main elements in the framework werealready being implemented in their most recents u r veys. How eve r, the report also clearly show sthat much still needs to be done to develop andharmonise statistics on all components of employ-ment in the informal economy.

Women and Men in the Informal Economy wa swritten by a team of consultants working with theTask Force and team members of the InternationalLabour Office responsible for preparation of thereport on Decent Work and the Informal Economy.The co-authors of the report and co-directors of theproject that produced it were Martha Chen,

H a r vard Unive r s i t y, and Joann Vanek, recentlyretired from the United Nations Statistics Div i s i o n .Jacques Charmes, University of Ve r s a i l l e s ,prepared the international data on employment inand contribution of the informal economy.M a rgarita Guerrero, United Nations StatisticalInstitute for Asia and the Pacific, provided statis-tical advice in all aspects of the project includingtable design and preparation of the final tables.Françoise Carré, Radcliffe Institute for AdvancedS t u d y, and Joaquín Herranz, MassachusettsInstitute of Te c h n o l o g y, prepared the report onnon-standard work in developed countries. T h ecountry case studies were prepared by DebbieB u d l e n d e r, the Community A g e n cy for SocialE n q u i r y, and Peter Buwembo and NoziphoShabala, Statistics South Africa (South A f r i c a ) ;Rodrigo Negrete, National Institute of Statistics,G e o g r a p hy and Information and MercedesPedrero, National Autonomous University ofM exico (Mex i c o ) ; and Jeemol Unni, GujaratInstitute of Development Research (India).Technical advice was also provided by RichardA n ke r, Ralf Hussmanns, Eivind Hoffmann andAdriana Mata of the ILO and Grace Bediako of theUnited Nations Statistics Division. Productionassistance was provided by Joanna Jackson, ILO,and Marais Canali, Harvard Unive r s i t y ; anda d m i n i s t r a t ive assistance was provided bySasithorn Santiwongsakul, ILO.

This report was prepared under the supervisionof Lin Lean Lim and funded by the GenderPromotion Programme (GENPROM).

PREFACE

List of Tables and Boxes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

Main Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

The Informal Economy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9Informal workers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9The Informal Sector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9The Informal Economy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11Why Improve Statistics on the Informal Economy? . . . . . . . . . . . . . . . . . . . . . . 13User-Producer Collaboration to Improve Statistics on the Informal Economy . . . . . . . . 14

Statistics on the Informal Economy . . . . . . . . . . . . . . . . . . . . . . . . . . . 17Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17Developing Countries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

Informal Employment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17Self-employment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22Informal Sector and Gross Domestic Product . . . . . . . . . . . . . . . . . . . . . . . 24

The Informal Economy in Developed Countries . . . . . . . . . . . . . . . . . . . . . . . . 26Part-time Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27Temporary Employment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28Self-employment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29

Illustrative Case Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33Country Case studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33

India . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33Mexico . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36South Africa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40

Sub-Group Case Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43Home-Based Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42Street Vendors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49

Improved Statistics for Policy Making . . . . . . . . . . . . . . . . . . . . . . . . . 57Why are statistics crucial ? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57Objectives for Future Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58Plan of Action . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58

Annexes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60

TABLE OF CONTENTS

Chapter 1Box: 1993 ICLS definition of the informal sector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11Box: A conceptual framework: the informal economy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

Chapter 2Developing countries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

Box: Estimating informal employment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18Box: Estimating informal sector contribution to GDP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25Table 2.1: Informal employment in non-agricultural employment, by sex 1994/2000 . . . . . . . . . . . . . . . 19Table 2.2: Wage and self-employment in non-agricultural informal employment, by sex 1994/2000 . . . . . . . 20Table 2.3: Components of informal employment as percentage of non-agricultural employment, 1997/2000 . . . 21Table 2.4: Women’s share of employment in the informal economy, 1997/2000 . . . . . . . . . . . . . . . . . . 22Table 2.5: Self-employment as percentage of non-agricultural employment . . . . . . . . . . . . . . . . . . . . 22Table 2.6: Share of women and men in non-agricultural self-employment . . . . . . . . . . . . . . . . . . . . . 23Table 2.7: Percentage distribution of non-agricultural self-employment by sector, 1990/2000 . . . . . . . . . . . 23Table 2.8: Contribution of informal sector to GDP in selected developing countries . . . . . . . . . . . . . . . . 24Table 2.9: Contribution of women and men in non-agricultural informal sector employment to GDP in selected

developing countries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

Developed countries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26Box Measuring employment in the informal economy in developed countries . . . . . . . . . . . . . . . . . . . 26Table 2.10: Part-time employment by sex, 1990/1998 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28

Chapter 3India . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33

Box: Estimating employment in the informal economy in India . . . . . . . . . . . . . . . . . . . . . . . . . . 35Table 3.1: Composition of employment in the informal economy in India, 2000 . . . . . . . . . . . . . . . . . . 34

Mexico . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36Box: Estimating informal employment and contribution of informal economy to GDP . . . . . . . . . . . . . . 37Table 3.2: Informal economy in Mexico, 1998: Employment and contribution to GDP . . . . . . . . . . . . . . 37Table 3.3: Composition of employment in the informal economy: Mexico . . . . . . . . . . . . . . . . . . . . . 38

South Africa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40Box: The formal and informal sectors as parts of a continuum . . . . . . . . . . . . . . . . . . . . . . . . . . . 41Box: Domestic workers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42Table 3.4: Composition of employment in the informal economy in South Africa, 2000 . . . . . . . . . . . . . . 40

Home-based workers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43Box: Self-employed, homeworkers, and employees . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45Box: Homeworkers in global value chains . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46Box: Recent shifts in homework in Chile . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47Box: Working at home in India . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48Table 3.5: Home-based workers in fourteen developing countries: Number, share of

non-agricultural employment, proportion women . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47

Street vendors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49Box: Typology of street vendors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50Box: Bellagio International Declaration of Street Vendors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50Box: Street vendors in Mexico City, Mexico . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51Box: Street vendors in Ahmedabad, India . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52Box: Street vendors in Durban, South Africa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53Table 3.6: Street vendors in ten developing countries: Estimated number, share of non-agricultural

employment, proportion women . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52Table 3.7: Size and contribution of informal sector in trade and women traders in informal trade . . . . . . . . . 53Table 3.8: Street food enterprises by women’s involvement (per cent of all units) . . . . . . . . . . . . . . . . . 54

Annex 1: Countries, areas and geographical groupings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60Annex 2: Self-employment as percentage of non-agricultural employment . . . . . . . . . . . . . . . . . . . . . 62

LIST OF TABLES AND BOXES

Size of the Informal Economy

■ Informal employment comprises one half tothree-quarters of non-agricultural employ m e n tin developing countries : specifi c a l l y, 48 percent of non-agricultural employment in NorthA f r i c a; 51 per cent in Latin A m e r i c a ; 65 percent in A s i a ; and 72 per cent in sub-SaharanAfrica. If South Africa is excluded, the shareof informal employment in non-agriculturale m p l oyment rises to 78 per cent in sub-SaharanAfrica. If data were available for additionalcountries in Southern Asia, the reg i o n a laverage for Asia would likely be much higher.

■ Some countries include informal employmentin agriculture in their estimates of informale m p l oyment. In these countries the inclusionof informal employment in agricultureincreases significantly the proportion ofinformal employ m e n t : from 83 per cent ofnon-agricultural employment to 93 percent oftotal employment in India; from 55 per centto 62 per cent in Mexico; and from 28 per centto 34 per cent in South Africa.

■ Three categories of non-standard or atypicalwork – self-employment, part-time work, andtemporary work – comprise 30 per cent ofoverall employment in 15 European countriesand 25 per cent of total employment in theUnited States. Although not all self-employ e d ,part-time workers, and temporary workers areinformally employed, the majority receive few(if any) employment-based benefits or protec-tion. In the United States, for instance, lessthan 20 per cent of regular part-time wo r ke r sh ave employ e r-sponsored health insurance orpensions.

Components of the InformalEconomy

• Informal employment is comprised of bothself-employment in informal enterprises (i.e.,small and/or unregistered) and wage employ-

ment in informal jobs (i.e., without securecontracts, wo r ker benefits, or social protec-tion). In all developing regions, self-employ-ment comprises a greater share of informale m p l oyment (outside of agriculture) than wa g ee m p l oy m e n t : specifi c a l l y, self-employ m e n trepresents 70 per cent of informal employ m e n tin sub-Saharan Africa, 62 per cent in NorthAfrica, 60 per cent in Latin America, and59 per cent in Asia. Excluding South Africa,where black-owned businesses were prohibitedduring the apartheid era and have only recentlybegun to be recognised and reported, the shareof self-employment in informal employ m e n tincreases to 81 per cent in sub-Saharan Africa.

■ S e l f - e m p l oyment represents nearly one-thirdof t o t a l non-agricultural employment wo r l d-wide. It is less important in developed coun-tries (12 percent of total non-agriculturale m p l oyment) than in developing countrieswhere it comprises as much as 53 per cent ofnon-agricultural employment in sub-SaharanAfrica, 44 per cent in Latin America, 32 percent in Asia, and 31 per cent in North Africa.

■ Informal wage employment is also significantin the developing wo r l d : comprising 30 to4 0 per cent of informal employment (outsideof agriculture). Informal wage employment iscomprised of employees of informal enter-prises as well as various types of informalwage wo r kers who work for formal enter-prises, households, or who have no fi xe demployer. These include casual day labourers,domestic wo r kers, industrial outwo r ke r s(notably homewo r kers), undeclared wo r ke r s ,and part-time or temporary wo r kers withoutsecure contracts, wo r ker benefits, or socialprotection.

■ Non-standard wage employment, much ofwhich is informal, is significant in the deve l-oped world. In 1998, part-time work repre-sented 14 per cent of total employment for theOrganization for Economic Co-operation andD evelopment (OECD) countries as a wholeand more than 20 per cent of total employ m e n t

MAIN FINDINGS

in eight of these countries. In the countries ofthe European Union, temporary wo r kcomprises 11 per cent of total employ m e n t .Comparable data on other categories of non-standard wage work that are even more likelyto be informal in nature – namely, contractwork, industrial outwork, and casual daylabour - are not available.

■ Home-based wo r kers and street vendors aret wo of the largest sub-groups of the informalwo r k f o r c e : with home-based wo r kers themore numerous but street vendors the morevisible of the two. Taken together they repre-sent an estimated 10-25 per cent of the non-agricultural workforce in developing countriesand over 5 per cent of the total workforce indeveloped countries.

Women and Men in the InformalEconomy

■ Informal employment is generally a larg e rsource of employment for women than for menin the developing world. Other than in NorthAfrica where 43 per cent of women wo r ke r sare in informal employment, 60 per cent ormore of women wo r kers in the deve l o p i n gworld are in informal employment (outside

agriculture). In sub-Saharan Africa, 84 percent of women non-agricultural wo r kers areinformally employed compared to 63 per centof male non-agricultural wo r ke r s ; in LatinAmerica 58 per cent for women in comparisonto 48 per cent for men. In Asia, the proportionof women and men non-agricultural wo r ke r sin informal employment is roughly equiva l e n t .

■ Although wo m e n ’s labour force participationrates are lower than men’s, the limited dataavailable point to the importance of women inhome-based work and street vending in devel-oping countries : 30-90 per cent of streetvendors (except in societies that restrictwo m e n ’s mobility) ; 35-80 per cent of a l lhome-based wo r kers (including both self-employed and homeworkers); and 80 per centor more of homewo r kers (industrialoutworkers who work at home).

■ Although wo m e n ’s labour force participationrates are lower than men’s, women representthe vast majority of part-time wo r kers inm a ny d eveloped countries. In 1998, wo m e ncomprised 60 per cent or more of part-timeworkers in all OECD countries reporting data.Women’s share of part-time work for specificcountries was as high as 98 per cent in Sweden,80 per cent in the United Kingdom, and 68 percent in both Japan and the United States.

Women and Men in the Informal Economy: A Statistical Picture8

Informal Workers

Street vendors in Mexico City; push-cart vendorsin New York city ; rickshaw pullers in Calcutta ;j e e p n ey drivers in Manila ; garbage collectors inB o g o t á ; and roadside barbers in Durban. T h o s ewho work on the streets or in the open-air are themore visible occupational groups in the informaleconomy. The streets of cities, towns, and villagesin most developing countries – and in many devel-oped countries - are lined by barbers, cobblers,garbage collectors, waste recyclers, and vendors ofvegetables, fruit, meat, fish, snack–foods, and amyriad of non–perishable items ranging from locksand keys to soaps and detergents, to clothing. Inm a ny countries, head–loaders, cart pullers, bicy c l epeddlers, rickshaw pullers, and camel, bullock, orhorse cart drivers jostle to make their way dow nnarrow village lanes or through the maze of cars,trucks, vans, and buses on city streets.

But the informal economy also includes activ-ities and wo r kers that are less visible and, eve n ,i nvisible. Less visible informal wo r kers work insmall shops and workshops. On the street cornersof most cities, towns, or villages, even in residen-tial areas, are countless small kiosks or stalls thatsell goods of every conceivable kind. Down thec r owded lanes of most cities, towns, or villages aresmall workshops that repair bicycles and motorcy-c l e s; recycle scrap metal; make furniture and metalp a r t s; tan leather and stitch shoes; weave, dye, andprint cloth; polish diamonds and other gems; makeand embroider ga r m e n t s; sort and sell cloth, paper,and metal waste; and more.

The least visible informal wo r kers, themajority of them women, sell or produce goodsfrom their homes: garment make r s ; embroiderers;incense–stick rollers; cigarette–rollers; paper bagm a ke r s ; kite make r s ; hair band make r s ; foodp r o c e s s o r s; and others. These least visible wo r ke r sare not confined to developing countries. Home-based wo r kers are to be found around the wo r l d ,i n c l u d i n g : garment wo r kers in To r o n t o ; embroi-derers on the island of Madeira ; shoemakers inM a d r i d ; and assemblers of electronic parts in

Leeds. Some of these home-based wo r kers wo r kon their own account, while others work on apiece-rate basis for a contractor or a firm.

Other categories of informal work are commonin both developed and developing countries: casualworkers in restaurants and hotels; sub-contractedjanitors and security guards ; casual or daylabourers in construction and agriculture; piece-rate workers in sweatshops; and temporary officehelpers or off-site data processors. Most wo r ke r sin all of these categories of work are informallyemployed, without secure contracts, worker bene-fits, or social protection.

Conditions of work and the level of earningsdiffer markedly among those who scavenge on thestreets for rags and paper, those who producegarments on a sub–contract from their homes,those who sell goods on the streets, and those whowork as temporary data processors. Even withincountries, the informal economy is highlys egmented by location of work, sector of thee c o n o m y, and status of employment and, acrossthese segments, by social group and gender. Butmost workers in the informal economy share onething in common: the lack of formal labour andsocial protection.

The Informal Sector

Discovery of the Informal Sector

It was widely assumed during the 1950s and 1960sthat, with the right mix of economic policies andresources, poor traditional economies could betransformed into dynamic modern economies. Inthe process, the traditional sector comprised ofpetty traders, small producers, and a range ofcasual jobs would be absorbed into the moderncapitalist – or formal – economy and, thereby,disappear. This perspective was reinforced by thesuccessful rebuilding of Europe and Japan afterWorld War II and the expansion of mass produc-tion in Europe and North America during the 1950sand 1960s. By the mid-1960s, how eve r, the opti-mism about the prospects for economic grow t h

1. THE INFORMAL ECONOMY

i n d eveloping countries began to give way toconcerns about persistent widespread unemploy-ment. Reflecting this concern, the InternationalLabour Organization (ILO) mounted a series ofl a rge multi-disciplinary “e m p l oyment missions”to various developing countries. The first of thesewas to Kenya in 1972.

The Ke nya employment mission, through itsfieldwork and in its official report, recognised thatthe traditional sector had not just persisted bu texpanded to include profitable and efficient enter-prises as well as marginal activities.1 To highlightthis fact, the Kenya mission chose to use the term“informal sector” rather than “traditional sector”for the range of small-scale and unreg i s t e r e deconomic activities. This term had been coined theyear before by a British economist, Keith Hart, inhis 1971 study of economic activities in urbanGhana. 2

Debates about the Informal Sector

Although both Hart and the Ke nya mission teamwere very positive about the informal sector –noting its effi c i e n cy, creativ i t y, and resilience – theconcept received a mixed rev i ew in deve l o p m e n tcircles. Many observers subscribed to the notionthat the informal sector was marginal or peripheraland not linked to the formal sector or to moderncapitalist development. Some of these observe r scontinued to believe that the informal sector inKe nya, Ghana, and other developing countrieswould disappear once these countries achieve ds u fficient levels of economic growth or modernindustrial development. Other observers arg u e dthat industrial development might take a differentpattern in developing countries – including theexpansion of informal economic activities - than ithad in developed countries.

By the 1980s, the focus of the informal sectordebate expanded to include changes that wereoccurring in advanced capitalist economies.I n c r e a s i n g l y, in both North America and Europe,production was being reorganised into small-scale,decentralised, and more flexible economic units.Mass production was giving way to “f l ex i b l es p e c i a l i s a t i o n” or, in some contexts, reverting tosweatshop production. 3 These new patterns ofcapitalist development were (and are still) associ-ated with the informalisation of employment rela-tions – standard jobs being turned into non-stan-dard or atypical jobs with hourly wages but fewb e n e fits or into piece-rate jobs with no benefits -and with sub-contracting the production of goodsand services to small-scale informal units andindustrial outworkers. In the process, the informal

economy becomes a permanent, albeit subordinateand dependent, feature of capitalist development. 4

Meanwhile, in the 1980s, the economic crisisin Latin America served to highlight anotherfeature of the informal sector : namely, thate m p l oyment in the informal sector tends to growduring periods of economic crisis. 5 In the A s i a neconomic crisis a decade or more later, millions ofpeople who lost formal jobs in the former EastAsian Tiger countries tried to find jobs or creatework in the informal economy. 6 Meanwhile, struc-tural adjustment in Africa and economic transitionin the former Soviet Union and in Central andEastern Europe were also associated with anexpansion of employment in the informal economy.W hy does employment in the informal economytend to expand during periods of economic adjust-ment or transition? When private firms or publicenterprises are downsized or closed, retrenchedwo r kers who do not find alternative formal jobsh ave to turn to the informal economy for wo r kbecause they cannot afford to be openly unem-ployed. Also, in response to inflation or cutbacksin public services, households often need to supple-ment formal sector incomes with informal earnings.

During the1990s, globalisation of the economyc o n t r i buted to the informalisation of the wo r k f o r c ein many industries and countries. 7 Whereas glob-alisation generates new jobs and new marke t s ,available evidence suggests that not all the jobs are“ g o o d ” jobs and that the most disadva n t a g e dproducers have not been able to seize new marketopportunities. This is because global competitionerodes employment relations by encouragingformal firms to hire wo r kers at low wages with fewb e n e fits or to sub-contract (or out-source) theproduction of goods and services8 ; and global inte-gration reduces the competitiveness of manyinformal firms or self-employed producers vis-à-vis imported goods (in domestic markets) and vis-à-vis larg e r, more formal firms (in export marke t s ) .

Statistics on the Informal Sector

At the 1991 International Labour Conference, theinformal sector was the topic for tripartite debate.This was the first time that the informal sector wasfeatured as a major agenda item in an internationalconference. 9 The 1991 tripartite debate reinforcedthe growing interest in statistics on the informalsector within the ILO Bureau of Statistics andamong the members of the InternationalConference of Labour Statisticians (ICLS). In1992, the ILO Bureau of Statistics prepared areport on “Statistics on Employment in theInformal Sector” as the basis for discussion at the


International Conference of Labour Statisticians(ICLS) meeting the following year. The 1993 ICLSadopted an international statistical definition of the“informal sector” that, subsequently, was includedin the revised International System of NationalAccounts (1993 SNA).

In order to be consistent with the framework ofthe System of National Accounts and provide for aseparate accounting of Gross Domestic Production(GDP) in the informal sector, the definition wa sbased on production units or enterprises rather thanon employment relations.1 0 The 1993 ICLS defi n e dthe informal sector as all unregistered or unincor-porated enterprises below a certain size, including: micro-enterprises owned by informal employerswho hire one or more employees on a continuingb a s i s; and own-account operations owned by indi-viduals who may employ contributing fa m i l ywo r kers and employees on an occasional basis( s e e box on 1993 ICLS definition of the info r m a lsector).

H ow eve r, the Conference of LabourStatisticians also recognised that an enterprise-based definition would not capture all dimensionsof informal employment. The Conference recom-mended that further work was needed on theemployment-based dimensions of informality. 11

The Informal Economy

The concept of the informal sector has beendebated since its “d i s c ove r y” in Africa in the early1970s. Nevertheless, it has continued to be used bym a ny policy makers, labour advocates, andresearchers because the reality it seeks to capture– the large share of the global workforce that

remains outside the world of full-time, stable, andprotected employment – continues to be importantand probably has been increasing over time. A tpresent, there is renewed interest in informal workarrangements. This current interest stems from thefact that informal work arrangements have not onlypersisted and expanded but have also emerged innew guises and unexpected places.

New Term and Expanded Concept

In recent years, some policy makers, activists, andresearchers - both within and outside the ILO –have started to use the term “informal economy”for a broader concept that incorporates certaintypes of informal employment that were notincluded in the 1993 international statistical defi-nition of the “informal sector”. T h ey seek to incor-porate in this concept the whole of informality -including both enterprise and employment rela-tions - as manifested in industrialised, transition,and developing economies.

This shift toward an expanded concept of the“informal economy” reflects a rethinking of someof the key assumptions regarding the so-called“ informal sector”. Those invo l ved in the currentrethinking, both within and outside the ILO, seekto incorporate the real world dynamics in labourm a r kets today – particularly the employ m e n tarrangements of low-income wo r kers. Keyelements of this alternative framework aresummarised below :■ What Is Not the Informal Economy:

H i s t o r i c a l l y, the informal sector was defi n e dl a rgely in opposition to the formal sector : asthe sum total of all income-earning activ i t i e soutside of legally regulated enterprises andemployment relations. But this is too inclusive

The Informal Economy 11

1993 ICLS Definition of the Informal SectorThe Resolution concerning Statistics of Employment in the Informal Sector, adopted by the

Fifteenth International Conference of Labour Statisticians in 1993, defined the informal sector asfollows: The informal sector is regarded as a group of household enterprises or unincorporatedenterprises owned by households that includes:• i n f o rmal own-account enterprises, which may employ contributing family workers and

employees on an occasional basis; and • enterprises of informal employers, which employ one or more employees on a continuous basis.

The enterprise of informal employers must fulfil one or both of the following criteria: size ofunit below a specified level of employment, and non-registration of the enterprise or its employees.

This framework proposed by the ICLS Resolution allows countries to adapt the basic opera-tional definition and criteria to their specific circumstances. In particular, flexibility is allowed withrespect to the upper limit on the size of employment; the introduction of additional criteria such asnon-registration of either the enterprise or its employees; the inclusion or exclusion of professionalsor domestic employees; and the inclusion or exclusion of agriculture.Source : ILO Report of the Fifteenth International Conference of Labour Statisticians, Geneva 1993.

a definition. As defined and used here, theinformal economy is distinguished from thefollowing :– Formal Economy: By definition, the

informal economy is understood to beoutside the world of reg u l a r, stable, andprotected employment and of legally regu-lated enterprises.

– Criminal Economy: The informal economyshould not be confused or conflated with thecriminal economy. While production ore m p l oyment arrangements in the informaleconomy are often semi-legal or illegal, theinformal economy produces and distributesl egal goods and services. In contrast, thecriminal economy deals in illegal goods andservices.

– Reproductive or Care Economy: The repro-d u c t ive or care economy – comprised ofunpaid domestic work and care activities –is also not part of the informal economy.This is because the informal economy isdefined as part of the market economy: thatis, as producing goods and services for saleor some other form of remuneration.

■ What I s the Informal Economy: In theexpanded conceptual framework the informaleconomy is seen as comprised of informale m p l oyment (without secure contracts, wo r ke rbenefits, or social protection) both inside andoutside informal enterprises (see box onA Conceptual Framewo r k: The Info r m a lEconomy) :– Informal Employment in Informal

Enterprises (small unregistered or unincor-porated enterprises), including: employers,e m p l oyees, own account operators, andunpaid family wo r kers in informal enter-prises.12

– Informal Employment outside InformalEnterprises (for formal enterprises, forhouseholds, or with no fi xed employ e r ) ,including: domestic workers, casual or daylabourers, temporary or part-time wo r ke r s ,1 3

industrial outwo r kers (including home-wo r kers), and unregistered or undeclaredworkers.

■ What Characterises Informal Employment:Under the expanded concept, informal employ-ment is understood to include all remunerativework – both self-employment and wa g ee m p l oyment - that is not recognised, reg u l a t e d ,or protected by existing legal or reg u l a t o r yf r a m eworks and non-remunerative wo r ku n d e r t a ken in an income-producing enterprise.

Most informal workers – including both self-employed and wage workers – are deprived ofsecure work, worker’s benefits, social protec-tion, and representation or voice. The self-employed have to take care of themselves andtheir enterprises. Moreove r, they often face ac o m p e t i t ive disadvantage vis-à-vis larg e rformal firms in capital and product marke t s .Informal wage workers also have to take careof themselves as they receive few (if any )employer-sponsored benefits. Moreover, bothgroups receive little (if any) legal or socialprotection. As a result of these and otherfactors, a higher percentage of people wo r k i n gin the informal economy, compared to thoseworking in the formal economy, are poor.

■ Continuum of Economic Relations:H i s t o r i c a l l y, many observers considered theinformal and formal sectors as two distincteconomic sectors without direct links to onea n o t h e r. The reality is more complex. To beg i nwith, production, distribution, and employ m e n trelations tend to fall at some point on acontinuum between “ f o r m a l” relations (i.e.,r egulated and protected) at one pole and“ i n f o r m a l” relations (i.e., unregulated andunprotected) at the other. Moreover, the moreformal and the more informal parts of theeconomy are often dynamically linked. Fo rinstance, many informal enterprises haveproduction or distribution relations with formalenterprises: supplying inputs, finished goods,or services to each other either through directtransactions or sub-contracting arrangements.Also, many formal enterprises hire wa g ewo r kers under informal employment relations:for example, many part-time workers, tempo-rary workers, and industrial outworkers workfor formal enterprises through contracting orsub-contracting arrangements.

■ Segmentation of the Informal Economy :The informal economy consists of a widerange of informal enterprises and informaljobs. Despite its heterogeneity, there are mean-ingful ways to classify its components egments, including: by type of economic unitand by employment status. Informal enter-prises consist of micro-enterprises (with anemployer plus some employees), family busi-nesses (with an owner operator and, some-times, unpaid family wo r kers), and ow naccount operations (with an individual owneroperator). Informal employment relationsconsist of employees of informal enterprises aswell as domestic wo r kers without a regular


contract, casual day labourers without a fi xe de m p l oy e r, temporary wo r kers who get wo r kthrough an agency, part-time wo r kers for a fi xe de m p l oy e r, industrial outwo r kers for formal orinformal firms (and their intermediaries), andunregistered or undeclared workers.

Why Improve Statistics on theInformal Economy?

Changing employment arrangements – associatedwith economic growth, transition or crisis and withglobal integration and competition – require notonly new ways of conceptualising the informal

economy but also new approaches to measuring itssize and contribution and to classifying those whowork in it. How can one predict and modeleconomic performance if a large share of totaloutput is not adequately measured and valued? Orh ow can one predict and model labour marke tb e h aviour if a major segment of the total wo r k f o r c eis not adequately measured or understood? Labourstatistics and national accounts need to moreadequately capture and measure the size andc o n t r i bution of the total economy, including theinformal economy.

Current debates on poverty focus largely onhuman capabilities to the relative neglect of marke topportunities. How can one understand and


A Conceptual Framework: The Informal Economy

(a) As defined by the Fifteenth International Conference of Labour Statisticians in 1993 (see box on 1993 ICLS definitionof informal sector)

(b) Households producing goods for their own final use and households employing domestic workers.

Dark blue cells refer to jobs that by definition do not exist in the type of production unit in question.Light blue cells refer to jobs which exist in the type of production unit in question but which are notrelevant to our concerns.The unshaded cells are the focus of our concern – they refer to types of jobs that re p resent thedifferent segments of the informal economy.Cells 1 and 5 : Contributing family workers: no contract of employment and no legal or socialp rotection arising from the job, in formal enterprises (cell 1) or informal enterprises (cell 5).(Contributing family workers with a contract of employment, wage, social protection, etc. wouldbe considered employees in formal employment).Cells 2 and 6: Employees who have informal jobs whether employed by formal enterprises (cell 2) or informal enterprises (cell 6). Cells 3 and 4: Own-account workers (cell 3) and employers (cell 4) who have their own inform a lenterprises. The informal nature of their jobs follows directly from the characteristics of the enter-prise they own.Cell 7 : Employees working in informal enterprises but having formal jobs. (This may occur, forexample, when enterprises are defined as informal using size as the only criterion).Cell 8: Members of informal producers’ cooperatives.Cell 9: Producers of goods for own final use by their household (e.g. subsistence farming).Cell 10: Paid domestic workers employed by households in informal jobs.S o u rc e: ILO, Decent Work and the Informal Economy R e p o rt VI, International Labour Conference, 90th Session, (Geneva 2002).

Jobs by status in employmentProduction Contributing Membersunits by Own-account Employers family Employees of producers’

type workers workers cooperativesInformal Formal Informal Formal Informal Informal Formal Informal Formal

Formalsector 1 2

enterprisesInformalsector 3 4 5 6 7 8

enterprises (a)

Households (b) 9 10

address poverty without having better statistics onthe income sources, the income levels, and thefinancial risks of those who are poor? There is alink – although not a complete overlap – betweenworking in the informal economy and being poor.This is because those who work in the informaleconomy are generally not covered by labour leg i s-lation or social protection and earn less, onaverage, than wo r kers in the formal economy.However, there is no simple relationship betweenworking in the informal economy and being pooror working in the formal economy and escapingpoverty. In many countries, male owner operatorsof micro-enterprises earn more, on average, thanl ow-skilled wo r kers in the formal economy. Inmost countries, women homewo r kers earn wa yb e l ow the minimum wa g e . 1 4 The link betweenworking in the informal economy and being poor,especially in the lowest-return activities, is strongerfor women than for men.

To focus the attention of economists and policym a kers on the links between informality andg r owth, statistics on the size and contribution ofthe informal economy are needed. To focus theattention of policy makers and academics one m p l oyment opportunities and employ m e n tarrangements as a key determinant of pove r t y,statistics on the links between informality, pove r t y,and gender are also needed.

User-Producer Collaboration toImprove Statistics on the InformalEconomy

Under the international statistical system, directedby the United Nations Statistical Commission, theInternational Labour Office has responsibility forsetting standards relating to the collection of labourforce statistics and for compiling and publishingo fficial labour force statistics from around theworld. As part of this mandate, the ILO’s Bureauof Statistics has played a lead role in deve l o p i n gmethods for the collection of data on employmentin the informal sector, in compiling and publishingo fficial statistics in this area, and in providing tech-nical assistance to national statistical offices toi m p r ove their data collection on employment in theinformal sector.

During the late 1990s, activists orga n i s i n ginformal wo r kers began collaborating withresearchers and statisticians to improve statisticson the informal economy. In lobbying for appro-priate policies, activists need improved statisticson the informal economy as a whole and on

s p e c i fic categories of informal wo r kers. Fo rinstance, in 1995, the Self-Employed Wo m e n ’sAssociation (SEWA) of India and the internationalalliance of home-based wo r kers (called HomeNet)needed statistics on homework worldwide for usein their lobbying efforts for an internationalc o nvention on homework. The researchers, whowere asked to collect statistics on homewo r k ,found that only seven countries in the world –Morocco, Japan, and five countries in Europe – hadthe category of homework in their official labourforce statistics. 15 Because so few official statisticswere available, the researchers collected and anal-ysed available data from micro-surveys around thewo r l d . 1 6 The statistics they collected helped toc o nvince government and wo r ker delegations tothe International Labour Conference that home-wo r kers were a significant part of the wo r k f o r c e ,especially in key export industries. The HomeWork Convention (No. 177) was passed in 1996.

In 1997, recognising the importance of thejoint action of activists, researchers, and statisti-cians, representatives of SEWA and HomeNetjoined other experts on the informal economyto form a global network called Women in Infor-mal Employ m e n t : Globalizing and Orga n i z i n g(WIEGO). One of W I E G O ’s five programmethemes is to improve and develop statistics on theinformal economy. In implementing this objective ,the statisticians and researchers in the W I E G On e t work work closely with the InternationalLabour Organization, the United Nations StatisticsD ivision and the International Expert Group onInformal Sector Statistics, known as the DelhiGroup.

That same year, SEWA started a collaborativeresearch project on the informal economy in Indiawith the National Council of Applied EconomicsResearch (NCAER) and the Gujarat Institute ofDevelopment Research. Among other studies, thisproject included a study on the size and contribu-tion of the informal economy in Ahmedabad citythat compared official data with the results of aspecial sample survey. Convinced by this study ofthe desirability and feasibility of better capturingthe informal economy in national sample surveys,the National Sample Survey Office (NSSO) of theGovernment of India commissioned the principleinvestigator of the Ahmedabad study to assist theNSSO to design special modules to measure therural and urban informal economy. These moduleswere administered in the 1999-2000 round of thenational sample survey; the findings are reportedin the case study of India (see Chapter 3).


During the late 1990s, other countries alsoundertook major new initiatives to collect data onthe informal sector and the informal economy. Forexample, South Africa and Mexico - the two othercountries featured as case studies in Chapter 3 -encouraged strong user- producer collaborations tod evelop new surveys to measure the informaleconomy. In the case of South Africa, the effort tobetter measure the informal economy was part ofa larger effort to develop the post-apartheidnational statistical system. In the case of Mexico,the effort to better measure the informal economywas part of a larger effort to monitor the impacto f the North America Free Trade A g r e e m e n t( NA F TA) and on-going privatisation of theeconomy.

Meanwhile, in 1997, the UN StatisticsCommission formed an International Expert Groupon Informal Sector Statistics, called the DelhiGroup. The convenors of the Delhi Group haveinvited the stakeholders in these various initiatives

– including the user groups - to participate in theDelhi Group meetings. These meetings haveprovided an important international forum for thediscussion and dissemination of the innova t ivemethods and new data coming out of these initia-tives. In collaboration with the ILO, the membersof the Delhi Group have encouraged the rethinkingof the informal economy.

This booklet reflects the joint action ofa c t ivists, researchers, and statisticians at thenational and international levels to improve statis-tics on the informal economy. All of the contribu-tors to this booklet have worked with one or moreof these initiatives. For the first time, the key find-ings from these and other recent initiatives arebeing presented in a single publication whichp r ovides a statistical picture of the informaleconomy worldwide. It is only a beg i n n i n g ,h ow eve r. As detailed in Chapter 4, a great dealmore work is needed to generate the data requiredfor improving policies on the informal economy.


NOTES1 International Labour Office, Employment, Incomes and Equality : A Strategy for Increasing Productive Employment in Kenya (Geneva, 1972).2 Keith Hart, “Informal Income Opportunities and Urban Employment in Ghana”, The Journal of Modern African Studies, Vol. 11, No. 1, pp.61-89 (1973). 3 Michael Piore and Charles Sabel, The Second Industrial Divide (New York, Basic Books, 1984).4 Alejandro Portes, Manuel Castells, and Lauren A. Benton, eds., The Informal Economy: Studies in Advanced and Less Developed Countries(Baltimore, John Hopkins University Press, 1989).5 Victor Tokman, ed., Beyond Regulation : The Informal Economy in Latin America (Boulder, Colorado, Lynne Rienner Publishers, 1992). 6 Eddy Lee, The Asian Financial Crisis: The Challenge for Social Policy (Geneva, International Labour Organization,1998). 7 Guy Standing, Global Labour Flexibility: Seeking Distributive Justice (New York, St. Martin's Press, 1999). 8 Dani Rodrik, Has Globalization Gone too Far? (Washington, D.C., Institute for International Economics, 1997). 9 Paul E. Bangasser, The ILO and the Informal Sector: An Institutional History, ILO Employment Paper, No. 2000/9.10 Ralf Hussmanns, Informal Sector and Informal Employment: Elements of a Conceptual Framework. Paper presented at the Fifth Meeting ofthe Expert Group on Informal Sector Statistics, Delhi (September 2001).11 Ibid.12 In the earlier conceptualization of the informal sector, informality was seen as a characteristic of enterprises not of employment relations. Forthis reason, the 1993 ICLS definition of the informal sector includes all employment, both formal and informal, inside informal enterprises.Formal employment within informal enterprises is a relatively rare phenomenon. In line with the expanded concept of informal economy, theterm “informal enterprises” is used in this report instead of “informal sector” except in two sections. The term “informal sector” is used in thesection on contribution to GDP in Chapter 2A in conformity with the framework used to distinguish the contribution of the informal sectorwithin the household sector; and in Chapter 3A because the three countries – India, Mexico, and South Africa – directly measured employmentin the informal sector.1 3 Temporary and part-time wo r kers who are covered by labour legislation and statutory social protection benefits are not included in the informaleconomy (see Chapter 2 Developed Countries for a discussion of the benefits of temporary and part-time workers).1 4 Jacques Charmes, Informal Sector, Poverty and Gender: A Rev i ew of Empirical Evidence. (Washington, D.C., The World Bank., 1998) ;S . V. Sethuraman, G e n d e r, Informality and Pov e r t y, (Washington, D.C., The World Bank, 1998); Martha Chen and Donald Snodgrass, ManagingResources, Activities, and Risk in Urban India: The Impact of SEWA Bank”, (Washington, D.C., USAID, 2001).15 G. Schneider de Villagas, “Home Work : A Case for Social Protection”, International Labour Review, Vol. 129, No. 4 (1990). 1 6 Martha Chen, Jennefer Sebstad, and Lesley O’Connell, “Counting the Invisible Wo r k f o r c e”, World Dev e l o p m e n t, Vol. 27, No. 3, pp. 603-610 (1999).

Introduction

Preparing the statistics for this chapter on the size,composition, and contribution of the informaleconomy was an extremely difficult exercise. Themost serious limitation is that very few countriesh ave undertaken regular surveys on the informalsector and only two or three countries havecollected the data that provide for measures ofinformal employment outside informal enterprises.Another limitation in the available data is that fewcountries have prepared estimates of the contribu-tion of the informal sector to GDP.

Second, there are a number of problems thatlimit the international comparability of the data. Inimplementing the international statistical defi n i-tion of the informal sector, countries applyd i fferent criteria of non-registration, enterprisesize, and/or workplace location. Also, according tothe 1993 ICLS guidelines, individual countries candecide what size of unregistered units to include inthe informal sector and whether (or not) to includeagriculture and domestic service in the informalsector.

F u r t h e r, the official data are not comprehen-s ive. Most countries exclude agriculture from theirmeasurement of the informal sector and somemeasure only the urban informal sector. Finally,the official data are not tabulated or presented in astandard format. Often it is not clear what the datainclude or exclude. Also, data on the informalsector (excluding agriculture) are often comparedto data on the total workforce (including agricul-ture), resulting in an under-estimation of the signif-icance of the informal sector.

In the absence of reliable data collectedd i r e c t l y, various indirect methods to estimate thesize and composition of the informal economyhave been used in this chapter.

The first part of this chapter presents estimatesof employment in the informal economy in devel-oping countries based on a residual method. It alsopresents data on self-employment worldwide as aproxy for informal self-employment, which is themajor component of informal employment. T h e

second part presents data on three categories ofnon-standard work in developed countries as aproxy for informal employment in deve l o p e dcountries.

Developing Countries

Informal Employment

The expanded conceptual framework on theinformal economy described in Chapter 1 has yetto be fully implemented in national data collectiona c t ivities. Notewo r t hy is the work undertaken ins everal countries that provides a basis for furthermethodological work (see the country casestudies in Chapter 3). In the absence of directmeasures, estimates were prepared using indirectmethods. Two types of estimates were prepared: a)estimates for 25 countries which have not carriedout an informal sector or mixed survey; and b) esti-mates for five countries that have carried out aninformal sector or mixed survey and therefore relyon indirect methods for only a part of the data–specifically on informal employment outside theinformal enterprises and total informal employ-ment (see box on estimating informal employ-ment by the residual method).1 In addition, directmeasures of self-employment, the major compo-nent of informal employment, are presented for70 developing and 30 developed countries.

According to the first set of estimates, in allregions of the developing world informal employ-ment (outside of agriculture) represents nearly halfor more of total non-agricultural employment. Itranges from 48 percent in North Africa, to 51 percent in Latin America, 65 percent in Asia and7 2 per cent in sub-Saharan Africa (table 2.1). Ifdata were available for Southern Asian countries,other than India, the regional average for A s i awould probably be higher still ( refer to annex 1for listing of countries in regional groups).

Informal activities in agriculture were notincluded in the estimates of employment in theinformal economy reported in this chapter. Since

2. STATISTICS ON THE INFORMAL ECONOMY


Estimating Informal EmploymentFew countries directly measure employment in the informal enterprises and fewer still measure

informal employment outside the informal enterprises. Until the requisite new data are collected,a l t e rnative methods for estimating informal employment are re q u i red. The so-called “re s i d u a lm e t h o d” is an indirect approach to estimating total informal employment and its major compo-nents outside of agriculture. It is based on existing published statistical data which are availablein many developing countries.

The basic data required for this estimation procedure are to be found in tabulations based onthe population census, labour force survey or another type of household survey that cross-classifyindustrial sectors (i.e., agriculture, mining, manufacturing, trade and services) by employment status(i.e., employers, own-account workers, family workers, employees) and by sex. While many coun-tries collect these data, they often do not tabulate or publish them. As a result, estimates of theinformal economy using the first residual method could be made for only 35 countries and fewer,a total of 25, had data disaggregated by sex. The second residual method re q u i res additionaldata on employment in the informal enterprises which is to be derived from special informal sectorsurveys or mixed (household and enterprise) surveys.

Residual Method 1The starting point is to determine the size of the total non-agricultural workforce .

The second step is to determine the number of formal employees in the non-agricultural work-force based on an establishment survey, an economic census, or other administrative sources thatrecord the legal status of firms. Those paid workers who work in corporations or quasi-corpora-tions or other legally re g i s t e red firms should be classified as formal employees. Generally, economiccensuses do not include civil servants and military forces. Therefore, it is also necessary to obtainan estimate of these government workers from other sources and add them to the number of formalemployees.

The third step is to estimate total informal employment (outside of agriculture) by subtractingthe number of formal employees (based on the economic census or establishment survey) from thetotal non-agricultural workforce (based on the population census or labour force survey). The esti-mates from the population census or labour force survey will always be larger than those from theeconomic census or establishment survey because the latter sources do not capture employmentoutside formal establishments. This diff e rence re p resents a residual estimate of total informal employ-ment.

The fourth step is to divide the total non-agricultural workforce into paid employees and self-employed . This can be done from the table classifying the workforce by employment status (i.e.,employers, own-account, contributing family workers and employees). This can be done for thenon-agricultural workforce only if the tabulations cross-classify employment by industrial sector andby employment status. An estimate can then be prepared for a) paid employees and b) the self-employed (employers, own account, and unpaid family workers) in the non-agricultural workforce.The self-employed so derived will typically include some high-end professionals and employers ofregistered enterprises, who are not considered to be informally employed. These categories areassumed to be small worldwide and smaller in developing countries than in developed countries.

The fifth step is to derive an estimate of informal paid employment (outside of agriculture) bysubtracting the number of self-employed (from step 4) from total informal employment (the residualobtained in step 3). It will include employees of informal enterprises. It may exclude some sub-contracted paid workers who declare themselves as self-employed.

Tables 2.1 and 2.2 are based on this residual method.

Residual Method 2 If a direct measure of employment in the informal enterprises is available (from mixed surveys orspecial informal sector surveys), then another residual method can be used. These kinds of esti-mates are available only for India, Kenya, Mexico, Tunisia, and South Africa.

The starting point is to determine the size of the total non-agricultural workforce . The second step is to determine the number of formal employees in the total non-agricultural

workforce, as above. The third step is to estimate total informal employment (outside of agriculture) by subtracting

the number of formal employees from the total non-agricultural workforce, as above. The fourth step is to determine total employment in the informal enterprises based on special

informal sector surveys or special mixed household-enterprise surveys. In some countries or indus-tries, it might include some formal or standard employees working in informal enterprises. But thiscategory is assumed to be very small. It might exclude some undeclared informal employees of

agricultural activities are an important source ofe m p l oyment, especially for women, their ex c l u s i o nreduces the overall estimate of the size of theinformal economy. Following national experienceand priorities, at least three countries – India,M exico, and South Africa – have distinguished

between formal and informal activities in agricul-ture and included informal agriculture (as a distinctc a t egory) in estimates of the informal economy.When informal employment in agriculture isadded, the significance of informal employ m e n tincreases (see three country cases in Chapter 3).

Statistics on the Informal Economy 19

informal enterprises. Although the latter category is assumed to be quite small, it is important tonote that both informal and formal enterprises may under-declare their paid employees.

The fifth step is to estimate i n f o rmal employment outside the informal enterprises . This is derivedby subtracting employment in the informal enterprises from the estimate of total informal employ-ment (the residual obtained in step 2). This second residual is an estimate of informal wage workoutside informal enterprises: that is, informal jobs for formal firms, for households, or when thereis no fixed employer. It may include some undeclared employees of informal enterprises. Also, itmay exclude some sub-contracted paid workers who declare themselves as self-employed.

Table 2.3 and 2.4 are based on this method.

Table 2.1 Informal employment in non-agricultural employment, by sex 1994/2000

North Africa 48 43 49Algeria 43 41 43Morocco 45 47 44Tunisia 50 39 53Egypt 55 46 57Sub-Saharan Africa 72 84 63Benin 93 97 87Chad 74 95 60Guinea 72 87 66Kenya 72 83 59South Africa 51 58 44Latin America 51 58 48Bolivia 63 74 55Brazil 60 67 55Chile 36 44 31Colombia 38 44 34Costa Rica 44 48 42El Salvador 57 69 46Guatemala 56 69 47Honduras 58 65 74Mexico 55 55 54Rep Dominicana 48 50 47Venezuela 47 47 47Asia 65 65 65India 83 86 83Indonesia 78 77 78Philippines 72 73 71Thailand 51 54 49Syria 42 35 43Source: Data prepared by Jacques Charmes. See also box on estimating informal employment.

Region/countryInformal employment as

percentage of non-agricultural employment

Women’s informal employment as percentageof women’s non-agricultural

employment

Men’s informal employmentas percentage of men’s non-

agricultural employment

Informal employment is generally a larg e rsource of employment for women than for men. Inmost countries – outside of North Africa – forwhich data are available, the proportions of wo m e nwo r kers in informal employment (outside ofagriculture) is larger than for men. Honduras,Indonesia and Syria are the exceptions. In thesecountries and in three of the four North A f r i c a ncountries, the proportions of men in informalemployment exceed those of women.

In some countries of sub-Saharan Africa mostof the female non-agricultural labour force is ininformal employment – 97 per cent in Benin,9 5 per cent in Chad, 87 per cent in Guinea and83 per cent in Kenya. Half or more of female non-agricultural wo r kers are in informal employ m e n t

in seven of the 11 Latin America countries and infour of the six Asian countries for which data areavailable. In India, 86 per cent of women in thenon-agricultural workforce are in the informaleconomy.

Informal employment – as estimated by thefirst residual method – is comprised of both selfand wage employ m e n t (table 2.2). In all reg i o n sself-employment is a larger share of non-agricul-tural informal employment than wage employ-ment. Self-employment is 70 per cent of informalemployment in sub-Saharan Africa, 62 per cent inNorth Africa, 60 per cent in Latin America, and5 9 per cent in Asia. Moreove r, across the indiv i d u a lcountries, self-employment is generally a greatershare of informal employment than wage employ-


Table 2.2 Wage and self-employment in non-agricultural informal employment, by sex 1994/2000

North Africa 62 72 60 38 28 40Algeria 67 81 64 33 19 36Morocco 81 89 78 19 11 22Tunisia 52 51 52 48 49 48Egypt 50 67 47 50 33 53Sub-Saharan Africa 70 71 70 30 29 30Benin 95 98 91 5 2 9Chad 93 99 86 7 1 14Guinea 95 98 94 5 2 6Kenya 42 33 56 58 67 44South Africa 25 27 23 75 73 77Latin America 60 58 61 40 42 39Bolivia 81 91 71 19 9 29Brazil 41 32 50 59 68 50Chile 52 39 64 48 61 36Colombia 38 36 40 62 64 60Costa Rica 55 49 59 45 51 41El Salvador 65 71 57 35 29 43Guatemala 60 65 55 40 35 45Honduras 72 77 65 28 23 35Mexico 54 53 54 46 47 46Rep Dominicana 74 63 80 26 37 20Venezuela 69 66 70 31 34 30Asia 59 63 55 41 37 45India 52 57 51 48 43 49Indonesia 63 70 59 37 30 41Philippines 48 63 36 52 37 64Thailand 66 68 64 34 32 36Syria 65 57 67 35 43 33Source: Data prepared by Jacques Charmes. See also box on estimating informal employment.

Country/RegionSelf-employment as percentage of

non-agricultural informal employment

Total Women Men

Wage employment as a percentage of non-agricultural informal employment

Total Women Men

ment (in 19 of the 25 countries and an identicalshare in one country). In one country – SouthAfrica – the difference is significant, in part due tothe history of apartheid with the prohibition ofb l a c k - owned businesses. The exclusion of SouthAfrica raises the share of self-employed in sub-Saharan Africa to 81 per cent.

In the countries reported in North Africa andAsia, in three of the five sub-Saharan Africa coun-tries and half of the Latin American countriess i g n i ficantly more women than men in informale m p l oyment (outside of agriculture) are self-e m p l oyed. By contrast, informal wage employ-ment is more important for women in Ke ny a ,South Africa and in four countries in SouthAmerica – Brazil, Chile, Columbia, and CostaRica. In these countries more than half of womenin informal employment are wage wo r ke r s .Moreover, in all but one of these countries, SouthAfrica, women are more likely to be wage wo r ke r sthan men.

For the second set of estimates, in countriesthat directly measure employment in informalenterprises (i.e., the informal sector) through eithera special informal sector survey or through mixed(household and establishment) surveys, theresidual method can be used to distinguish betweeninformal employment in informal enterprises andinformal paid employment outside informal enter-prises (see residual method 2 in box). In the pastfew years, five countries have collected such data.In all five countries, informal employment repre-sents at least half of non-agricultural employment(table 2.3) with rates of 50 per cent in Tu n i s i a ,51 per cent in South Africa, 55 per cent in Mexico,72 per cent in Kenya, and 83 per cent in India.

Both components of the informal economy –employment in informal enterprises and informalwage employment outside informal enterprises –

are important sources of non-agricultural employ-ment in these countries. In India, employment ininformal enterprises is especially signifi c a n t ; itcomprises 69 per cent of non-agricultural employ-ment; and informal employment outside informalenterprises is estimated at 14 per cent of non-agri-cultural employment. In total, nearly 85 per centof all non-agricultural employment in India is inthe informal economy. In South Africa, informalemployment outside informal enterprises is moreimportant as a source of employment than employ-ment in informal enterprises (32 per cent of totalnon-agricultural employment in comparison to1 9 per cent). For Mexico, Ke nya and Tunisia thet wo components tend to be more similar in size,each accounting for about one quarter to one thirdof total non-agricultural employment.

In the two sub-Saharan African countries,women are the majority of wo r kers in non-agri-cultural informal employment (table 2.4), specif-ically 60 per cent of workers in Kenya and 53 percent in South Africa. In Mexico women alsocomprise a large share of workers in non-agricul-tural informal employment (38 per cent). In Indiaand Tunisia women’s share is significant but muchsmaller (20 per cent and 18 per cent respectively).

In Kenya and South Africa, women’s share ofinformal employment outside informal enterprisesis especially significant. T h ey comprise 79 per centof these wo r kers in Ke nya and 61 per cent in SouthAfrica. In these two countries, women also repre-sent over 40 per cent of workers in informal enter-prises. In Mexico wo m e n ’s share in each of the twocomponents of informal employment is aboutequal. In India women’s share is the same in eachof the components; whereas in Tunisia wo m e n ’sshare of informal employment outside the informalenterprises (22 per cent) is higher than their sharein informal enterprises (15 per cent).


Table 2.3 Components of informal employment as percentage of non-agriculturalemployment, 1997/2000

India 83 69 15Mexico 55 29 26Kenya 72 36 35South Africa 51 19 32Tunisia 50 22 28Source: Data prepared by Jacques Charmes. See also box on estimating informal employment.

CountryTotal informal employment

as percentage of non-agricultural employment

Employment in informalenterprises as percentage

of non-agricultural employment

Informal employment outside of informal

enterprises as percentage ofnon-agricultural employment

Self-employment

M a ny of the wo r kers identified in Chapter 1- thoseengaged in petty trading, service repairs, transportand small manufacturing activities – are indepen-dent, self-employed producers, some of whomemploy family labour or a few additional workers.The most comprehensive set of statistics now ava i l-able on aspects of employment in the informaleconomy relates to these self-employed wo r ke r s .In the absence of direct measures of thosee m p l oyed in the informal economy, self-employ-ment provides an indicator of a major componentof employment in the informal economy. 2

Between 1980 and 2000 self-employ m e n tincreased from about one-quarter to about one-

third of non-agricultural employment wo r l d w i d e(table 2.5 and annex 2). Self-employment in non-agricultural activities increased in almost all deve l-oping regions. However, declines in self-employ-ment occurred in Eastern Europe where it droppedsharply, from 18 to 9 per cent of non-agriculturale m p l oy m e n t ; and in Eastern Asia, where itdropped from 23 to 18 per cent.

In the 1990 to 2000 period, in some deve l o p i n gr egions self-employment comprised half or moreof non-agricultural employment – notably in theCaribbean, Southern Asia and Sub-Saharan A f r i c a .With the exception of Eastern and Western A s i a ,where self-employment is 18 and 24 per cent,r e s p e c t ive l y, of the non-agricultural wo r k f o r c e ,s e l f - e m p l oyment is over 30 per cent of non-agri-


Table 2.4 Women’s share of employment in the informal economy, 1997/2000

India 20 20 20Mexico 39 40 38Kenya 60 42 79South Africa 53 45 61Tunisia 18 15 22Source: Data prepared by Jacques Charmes. See also box on estimating informal employment.

CountryWomen’s share of non-

agricultural employment inthe informal economy

Women’s share of non-agricultural employment in

informal enterprises

Women’s share of non-agricultural informal employment outside informal enterprises

Table 2.5 Self-employment as percentage of non-agricultural employment

WORLD 26 28 25 32 34 27Developed regions 13 11 14 12 10 14Eastern Europe 18 15 21 9 7 11Western Europe 12 10 14 14 11 17Other developed countries 10 9 11 11 9 12Africa 44 58 37 48 53 37Northern Africa 23 23 22 31 38 28Sub-Saharan Africa 50 69 41 53 57 40Latin America 29 30 27 44 54 35Central America 30 32 27 40 54 29South America 29 29 29 43 51 38Caribbean 27 28 25 55 67 43Asia 26 24 26 32 32 30Eastern Asia 23 20 24 18 26 16South-eastern Asia 34 38 31 33 40 24Southern Asia 40 38 40 50 55 48Western Asia 13 6 14 24 10 22Source: Data prepared by Jacques Charmes. See also technical note with Annex 2.

Region1980/1989

Total Women Men

1990/2000

Total Women Men

cultural employment in all sub-regions of thedeveloping world. Self-employment is less impor-tant in the developed regions where it is about12 per cent of non-agricultural employment.

S e l f - e m p l oyment is more important as a sourceof employment among women wo r kers thanamong men workers. In 1990/2000, in every sub-r egion of the developing world a greater propor-tion of women than men in non-agriculturale m p l oyment were self-employed (table 2.5). In

Africa and Latin America, roughly 50 per cent ofwomen in contrast to about 35 per cent of men ares e l f - e m p l oyed. How eve r, because many more menthan women are in the labour force, men comprisea larger share of self-employed non-agriculturalwo r kers than women wo r l d - w i d e : 64 per cent incontrast to 36 per cent (table 2.6). The only ex c e p-tion is Latin America where 51 per cent of self-e m p l oyed non-agricultural wo r kers are women and49 per cent are men.

By economic sector, outside of agriculture,s e l f - e m p l oyment is more important in trade thanin industry and services in both Africa and A s i a(table 2.7) : in Africa trade represents 47 per centof non-agricultural self-employment and in A s i a44 per cent. 3 By contrast, services are a moreimportant source of self-employment in LatinAmerica where services comprise 45 per cent ofs e l f - e m p l oyment – slightly higher in CentralAmerica and South America and signifi c a n t l ylower in the Caribbean.

Within the sub-regions there is more variationin the importance of the different economic sectorsfor wo m e n ’s self-employment than for men’s .Social norms that constrain women’s mobility areoften reflected in a relatively low incidence ofwomen street traders and a relatively high inci-dence of women working at home, particularly in


Table 2.6 Share of women and men innon-agricultural self-employment

WORLD 36 64

Developed regions 36 64

Africa 38 62Northern Africa 28 72Sub-Saharan Africa 41 59

Latin America 51 49

Asia 27 73Source: Data prepared by Jacques Charmes. See also technicalnote with Annex 2.

Region1990/2000

Women Men

Table 2.7 Percentage distribution of non-agricultural self-employment by sector:1990/2000

Region Industry Trade ServicesTotal Women Men Total Women Men Total Women Men

WORLD 30 25 33 38 39 35 32 36 33

Developed regions 33 19 37 31 41 29 36 40 34Eastern Europe 44 29 46 26 43 26 30 29 28Western Europe 27 14 32 36 44 32 37 41 36Other developed countries 28 15 35 29 33 26 43 52 39

Africa 32 29 32 47 51 45 21 20 23Northern Africa 33 52 20 40 26 52 28 22 28Sub-Saharan Africa 32 21 36 49 60 42 19 19 22

Latin America 24 14 32 31 27 27 45 58 40Central America 24 16 31 27 20 26 48 64 44South America 25 14 33 29 26 26 46 60 41Caribbean 19 8 35 48 51 35 33 41 30

Asia 27 41 28 44 33 42 29 25 32Eastern Asia 21 12 27 51 40 36 28 48 37South-eastern Asia 18 16 20 51 57 43 31 18 45Southern Asia 38 61 34 38 17 42 24 23 24Western Asia 27 42 27 42 28 43 31 30 30Source: Data prepared by Jacques Charmes. See also technical note attached to Annex 2.

m a n u facturing (see Chapter 3). In SouthernA s i a as a whole – including countries such asBangladesh, India, and Pakistan – 60 per cent ormore of the self-employed women in non-agricul-tural activities are in industry ; in Western A s i a4 2 per cent of the self-employed women are ini n d u s t r y ; but in Eastern and South-eastern A s i aless than 20 per cent are. Self-employment inindustry is also important for women in NorthernAfrica. Trade is the most important source ofe m p l oyment among self-employed women of sub-Saharan Africa (60 per cent of non-agriculturals e l f - e m p l oyment) and for women in the Caribbean(51 per cent). Services are important for self-employed women in Latin America, especially inCentral America (64 per cent) and South America(60 per cent) and also for self –employed womenin Eastern Asia (48 per cent).

Informal Sector and Gross DomesticProduct

There are, as yet, no estimates of the contributionof the informal economy as a whole to GrossDomestic Product (GDP).4 H ow eve r, there are esti-mates of the contribution of the informal sector toG D P. These estimates indicate that the contribu-tion of informal enterprises (i.e., the informalsector) to non-agricultural GDP is signifi c a n t(table 2.8). The average (unweighted) share of theinformal sector in non-agricultural GDP va r i e sfrom a low of 27 per cent in Northern Africa to ahigh of 41 per cent in sub-Saharan Africa. The fa c tthat a large number of countries in sub-SaharanAfrica have such estimates reflects a recognitionof the importance of the informal sector in totalG D P. The contribution of the informal sector toGDP is 29 percent for Latin America and 41 percent for Asia. How eve r, there is wide va r i a t i o namong the countries reported – from 13 per cent inMexico to 58 per cent in Ghana. This reflects notonly differences in the actual contribution of theinformal sector to GDP but also differences in themethods used among the countries in preparingthese estimates (see box on estimating informalsector contribution to GDP).

Data which permit analysis of the contributionof women and of men to the informal non-agricul-tural economy are available for only nine countries(table 2.9). Women’s contribution is significantlyhigher than men’s in Burkina Faso, Chad and Mali.The relatively high contribution of women inBurkina Faso has been explained by inclusion inthe official accounts of the contribution throughsecondary activities which for women are mainly

non-agricultural but for men are in agriculture.5 InBenin, the contribution of women and men arecomparable. In Kenya and Philippines, the contri-bution of women is lower than that of men; and, inTunisia, Indonesia and India, wo m e n ’s contribu-tion is substantially lower.


Table 2.8 Contribution of informal sectorto GDP in selected developing countries

Informal sector GDPCountry (year) as percentage of

non-agricultural GDP

AfricaNorthern Africa 27Algeria (1997) 26Morocco (1986) 31Tunisia (1995) 23Sub-Saharan Africa 41Benin (1993) 43Burkina Faso (1992) 36Burundi (1996) 44Cameroon (1995-96) 42Chad (1993) 45Côte d’Ivoire (1995) 30Ghana (1988) 58Guinea Bissau (1995) 30Kenya (1999) 25Mali (1989) 42Mozambique (1994) 39Niger (1995) 54Senegal (1991) 41Tanzania (1991) 43Togo (1995) 55Zambia (1998) 24

Latin America 29Colombia (1992) 25Mexico (1998) 13Peru (1979) 49

Asia 31India (1990-91) 45Indonesia (1998) 31Philippines (1995) 32Republic of Korea (1995) 17Source: Data prepared by Jacques Charmes. Also see box onestimating informal sector in GDP.


Estimating informal sector contribution to GDPAccording to the 1993 System of National Accounts (SNA), the informal sector is regarded

as a group of production units which form part of the household sector. However the whole contri-bution of the household institutional sector in the SNA cannot be attributed to the informal sector.P roduction for own final use is not included, which means that “ imputed re n t s ,” paid domestics e rvices and subsistence agriculture are excluded. The estimation of the informal sector in GDPrequires that the accounts of the household institutional sector are available by detailed industrialsectors, making it possible to distinguish agriculture and other primary activities as well as the esti-mates for rents and domestic service.

Although the 1993 SNA provided a framework for distinguishing the informal sector withinthe household sector, it does not provide specific guidelines for compiling independent estimatesof the informal sector. Table 2.8 presents experimental estimates of the contribution of the informalsector to GDP for a number of countries. Some of the estimates are official. Other estimates basedon official data differ from the official estimates either because activities which are to be excludedunder the 1993 SNA were removed from the official estimates or the more detailed informationused in preparing the official accounts were not available.

As shown in Table 2.8, Africa is the region where the greatest number of estimates are avail-able. In this region the informal sector was early assumed to represent an important share of totalp roduction as the formal sector was rather weak. Eff o rts were made to include estimates of thecontribution of the informal sector, often with the initial preparation of the national accounts in thec o u n t ry. In Asian countries the concern for measurement of the informal sector is more recent. India,Indonesia and the Philippines now make such estimates on a regular basis. However in LatinAmerica these estimates are generally not made. Mexico was the first Latin American country toimplement the new SNA and took the initiative of building a satellite account for the informal sector.Peru has estimated the contribution of individual entrepreneurs in the household institutional sectorwhich can serve as a proxy for the contribution of the informal sector to GDP. A tentative estimateis also available for urban Columbia.

Table 2.9 Contribution of women and men in non-agricultural informal sectoremployment to GDP in selected developing countries

AfricaBenin (1992) 43 22 21Burkina Faso (1992) 36 29 7Chad (1993) 45 28 17Kenya (1998) 25 11 16Mali (1989) 42 26 14Tunisia (1994-96) 23 4 19

AsiaIndia (1993) 47 10 18Indonesia (1998) 37 15 22Philippines (1995) 33 14 26Source : Data prepared by Jacques Charmes.

Country (year)Non-agricultural informal

sector GDP as percentage oftotal non-agricultural GDP

Women’s non-agriculturalinformal sector GDP

as percentage of total non-agricultural GDP

Men’s non-agriculturalinformal sector GDP as

percentage of total non-agricultural GDP


The Informal Economyin Developed Countries

With the expansion of standardized mass produc-tion and large corporations in Europe, Japan,Korea, and the United States in the post-World Wa rII period, it seemed as if most workers would havestandard wage jobs with employment-based bene-fits and social protection. How eve r, it is now recog-nized that an increasing share of the workforce ind eveloped countries works under non-standardarrangements, including part time and temporarywork and self-employment. As with informale m p l oyment in developing countries, many ofthese non-standard arrangements offer limitedworker benefits and social protection (see box onmeasuring employment in the info r m a leconomy in developed countries).

Non-standard employment arrangements ares i g n i ficant and even growing in developed coun-tries. In 1988, overall in fifteen countries of theEuropean Union (EU), non-standard arrangements– including part-time employment, temporaryemployment and self-employment in non-agricul-tural industries – represented one-quarter of totalemployment. By 1998, non-standard employmentin these countries had grown to nearly 30 percentof total employ m e n t . 6 Estimates for the UnitedStates also show about 25 per cent of the wo r k f o r c ein non-standard arrangements.7 This encompasseswo r kers employed under the following arrange-

m e n t s: part-time, temporary help agency, direct-hiretemporary (or fi xed term), on call, independentcontract, and contract firm wo r kers. While non-standard work increased significantly in the UnitedStates during the 1980s, the percent of workers innon-standard arrangements in the United States didnot increase between 1995 and 2001.

C o m p r e h e n s ive and comparable data on theextent of non-standard employment in deve l o p e dcountries is not now available. This is partly becaused e finitions of the various non-standard wo r karrangements are not comparable across countries.This is also because few countries collect data onall categories of non-standard work. For instance,there is only very limited data on some of the newforms of employment, such as employment bycompanies providing services to other companiesunder an inter- firm contract. Inter- firm subcon-tracting is a relatively new work arrangement thathas emerged in the fast growing sectors thatp r ovide labour- i n t e n s ive services to other bu s i-nesses or public institutions (e.g., janitorial services).

Assuming comprehensive data on non-stan-dard employment were available, there is thefurther question of which categories of non-stan-dard employment should be classified as informale m p l oyment. All categories of non-standardworkers would need to be classified by whether ornot they receive wo r ker benefits and labour andsocial protection coverage to arrive at the share ofnon-standard wo r kers who are informal wo r ke r s .

Measuring Employment in the Informal Economy in Developed CountriesEmployment in the informal economy in developed countries is linked to employment arrange-

ments that have tenuous connections between the worker and formal stru c t u res and little if anyaccess to fundamental labour rights and protection and employment-based social protection. Thefollowing specific categories of non-standard employment arrangements are relevant:• part-time employment in formal firms • temporary employment in formal firms• self-employment, in particular own account workers with no employees• inter-firm subcontracted work• industrial outwork, including homework• sweatshop work• day labour

Data on the first three categories are reviewed in this section to provide a proxy estimate ofthe size and composition of the informal economy in developed countries. This is because data oninter-firm subcontracting are available only for the United States and data on industrial outwork,sweatshop work, and day labour are not readily available.

To arrive at a more precise estimate of employment in the informal economy in developed coun-tries, two types of additional information are needed: data on all categories of non-standard work ;and data on access to worker benefits and social protection across all categories of non-standardwork. For instance, some part-time and temporary workers in developed countries enjoy workerbenefits and social protection coverage. Also, self-employed professionals and technicians indeveloped countries often can afford to subscribe to privatised social insurance schemes. Thosecategories of non-standard workers, such as these, who receive social protection coverage or canpay for social insurance would not be considered to be in informal employment.

Some temporary and part-time wo r kers are cove r e dby social protection and labour legislation whileothers are not ; this varies by country. Some self-employed are entitled to and can afford to pay fortheir own social protection (see discussion of bene-fits under each category of wo r ker below). A tpresent, data on certain categories – notably, part-time work, temporary work, and self-employment– can be compared cross nationally, as presentedb e l ow. Also, some data on the social protectionavailable to these workers is available, as analysedb e l ow. How eve r, comprehensive and systematicb r e a k d own of the non-standard workforce bywhether or not they receive wo r ker benefits andsocial protection coverage is not yet available.

Part-time Work

Since the beginning of the 1990s, part-time work– defined as less than 30 hours per week in a mainjob – has grown in most countries of theO rganization for Economic Co-Operation andDevelopment (OECD). Countries with the fastestg r owth in this period include France (from 12 to15 per cent), the United Kingdom (20 to 23 percent), Luxe m b o u rg (8 to 13 per cent) and Germany(13 to 17 per cent). By 1998, for the countries ofthe OECD as a whole, part-time employ m e n trepresented about 14 percent of total employment.Part-time work ranged from a high of 30 percentof total employment in the Netherlands to about 3per cent in the Czech Republic and Hungary. Part-time employment is also a high proportion of totale m p l oyment in Australia (26 per cent), Switzerland(24 per cent) Japan (24 per cent) and the UnitedKingdom (23 per cent) (table 2.10).

Women comprise the majority of part-timeworkers in each of the OECD countries. Women’sshare of part-time employment ranges from 60 percent in Turkey to 97 per cent in Sweden.

The main occupations of part-time wo r kers areservice and sales, clerical and low-skilled labour. 8

Women part-time wo r kers are concentrated inservice and sales (33 per cent of wo m e n ’s totalpart-time employment and 12 per cent for men)and clerical work (25 per cent for women and1 2 per cent for men). Men in part-time work areusually employed as low-skilled labourers (26 percent of men’s total part-time employment and23 per cent of women’s).

For most of the countries with data, the medianhourly earnings for part-time wo r kers are low e rthan those for full-time wo r kers in most sectors.Hourly earnings of part-time wo r kers representbetween 55 to 90 per cent of those of full-time

wo r ke r s .9 O verall the relative hourly earning ofpart-time workers are lowest in real estate, rentingand business activities with men earning less thant wo-thirds and women less than three-quarterso f the full-time wo r kers. This sector typicallye m p l oys 15 per cent or more of total part-timee m p l oyment. How eve r, in wholesale and retailtrade, which in most countries is the largest shareof part-time employment, the gap is much smaller.Women tend to earn more than 90 per cent of thehourly earnings of full-time wo r kers and menaverage just over 80 per cent. That is, the relativepay differential between female part-timers andfull timers is less than the equivalent pay differen-tial for males. One possible explanation for this isthat the full-time jobs of women are not thatd i fferent from the part-time jobs for wo m e n .Another is that male part-time jobs are limited toa small group of occupations associated with rela-tively low skills and earnings.

Benefits

As a general rule, part-time wo r kers tend to receivef ewer benefits than full-time wo r kers. How eve r,part-time workers in Europe are more likely to beeligible for benefits than part-time wo r kers inNorth America or Japan and Korea. In theEuropean Union, laws proscribe discrimination bye m p l oyers against part-time wo r kers in pay, certainb e n e fits and working conditions. 1 0 In additionm a ny European countries have collective agree-ments supporting the principle that part-timewo r kers are entitled to the same rights and bene-fits as full-time wo r ke r s . 1 1 In some countries,h ow eve r, these protections do not apply to part-time workers who work below a certain thresholdnumber of hours. For example, public health, old-age pension and unemployment benefits in France,G e r m a ny, Ireland, Japan and Sweden requireminimum hours or earnings for eligibility whichpart-time workers may not meet. 12

In Canada, part-time wo r kers are less like l ythan full-time wo r kers to be entitled to companypensions, health and dental plans, paid sick leaveand paid vacation leave. Data from the 1995 Surveyof Work Arrangements show that entitlements foreach of these benefits for part-time workers wereless than two - fifths of those for full- time wo r ke r s .1 3

In the United States, where part-time work isless common than in most other OECD countries,part-time wo r kers are offered very few benefi t s .Only 18 per cent of wo r kers in regular part-timejobs have employ e r-sponsored health insurance,compared to 87 per cent of regular full-timeworkers. Further, 19 per cent of regular part-time



wo r kers have employ e r- p r ovided pensions in com-parison to 60 per cent of regular full-time wo r ke r s .1 4

Japanese employers are not obligated to paysocial security, disability and unemployment insur-ance taxes for many part-time and temporaryworkers. 15 The situation is mitigated with regardsto health insurance, since Japan has a unive r s a lprogramme with everyone enrolled in some formof health insurance.

Temporary Employment

Temporary employment encompasses all short-term employment arrangements, whether throughan intermediary (e.g. a temp agency) or by directhire for a fixed term. More so than part-time workit is associated with few (if any) benefits.

Under a broad definition that includes tempo-rary agency wo r kers, wo r kers with fi xe d - t e r m

Table 2.10 Part-time Employment1 by sex, 1990-1998

Part-time employment Women’s share as a proportion in part-time

of total employment employment

1990 1998 1990 1998

Australia 2, 3 22.6 25.9 70.8 68.6Austria –.0 11.5 –.0 86.9Belgium 14.2 16.3 79.9 82.4Canada 17.0 18.7 70.1 69.5Czech Republic –.0 3.3 –.0 70.0Denmark 19.2 17.0 71.5 68.5Finland 7.6 9.7 66.8 63.1France 12.2 14.8 79.8 79.3Germany 13.4 16.6 89.7 84.1Greece 6.7 9.2 61.1 63.6Hungary –.0 3.4 –.0 69.2Iceland 4 –.0 23.2 81.6 77.4Ireland 9.8 – 71.8 –.0Italy 8.8 11.8 70.8 70.4Japan 2, 5 19.1 23.6 70.5 67.5Korea 2 4.5 6.8 58.7 54.8Luxembourg 7.6 12.8 86.5 87.3Netherlands 28.2 30.0 70.4 75.8New Zealand 19.6 22.8 77.1 74.3Norway 21.3 21.0 82.7 79.1Poland 2 –.0 11.8 –.0 62.2Portugal 6.8 9.9 74.0 70.9Spain 4.6 7.7 79.5 75.9Sweden 14.5 13.5 81.1 97.3Switzerland 4 22.1 24.2 82.4 83.4Turkey 9.2 6.2 62.5 60.3United Kingdom 20.1 23.0 85.1 80.4United States 6 13.8~ 13.4 68.2~ 68.0Total OECD 7 13.4~ 14.3 74.1~ 73.6Source: Data are published in OECD Employment Outlook (Paris: OECD, 1999) table E.~ Indicates break in series1 Part-time employment refers to persons who usually work less than 30 hours per week in their main job. Data include only personsdeclaring usual hours.2 Data are based on actual hours worked.3 Part-time employment based on hours worked at all jobs.4 1990 refers to 19915 Less than 35 hours per week.6 Estimates are for wage and salary workers only.7 For above countries only.

contracts, seasonal wo r kers, and persons withtraining contracts, most European Union countrieshave experienced increases in temporary employ-ment. For the European Union as a whole, theshare of wo r kers in temporary employ m e n tincreased between 1988 and 1998 from 8 to 11 percent of total employment.16

In many countries of the European Union themajority of workers in temporary employment arewomen. In nine of the fifteen European Unioncountries, women account for about half or moreof temporary employment. In six of these, womencomprised a solid majority: Sweden (59 per cent),Ireland (58 per cent), Belgium (57 per cent),United Kingdom (55 per cent), Netherlands (54 percent) and Denmark (52 per cent). The countries ofthe European Union with the lowest shares are :Spain (38 per cent), Austria (43 per cent) Greece(44 per cent) and Germany (45 per cent). 17

O verall for the 15 countries of the EuropeanUnion, temporary employment is more heav i l yconcentrated, than standard employment, inservice producing activities. This broad categ o r yincludes all services to individuals, households andbusinesses, including finance, insurance and realestate. Because temporary employment concen-trates in sectors that employ large proportions ofwomen wo r kers, women are heavily represented intemporary employment. 18

For the United States, temporary wo r ke r s(temporary agency, contract, on-call/day labourer,part-time, and full-time wo r kers) were about4 . 1 percent of all employed in 1999. 1 9 In Japantemporary employment (temporary, day labourers,and some temporary agency wo r kers) was about12 per cent of all employed in 1997 2 0 and themajority of these wo r kers were women (66 percent).

Benefits

European Union countries mandate parity in wa g e sand benefits for workers in most forms of tempo-rary employment, whether direct-hire or through atemporary agency.2 1 Although such wo r kers remaineligible for all socially administered employment-based benefits, they may not meet some of thehours, seniority, and earnings thresholds necessaryfor eligibility as their work is not stable.

In Canada, temporary or contract wo r kers wereless likely than full-time workers to be entitled toc o m p a ny pensions, health plans, dental plans, paidsick leave, and paid vacation leave. According tothe 1995 Survey of Work Arrangements, only20 per cent of non-permanent workers were enti-tled to an employer sponsored pension plan other

than the Canada/Quebec Pension Plan while 58 percent of full-time workers were entitled. 22

S i m i l a r l y, in the United States temporarywo r kers are much less likely than regular full-timewo r kers to have health insurance and priva t epension plans. In 2001, 11 per cent of temporaryhelp agency wo r kers and 30 per cent of on callwo r kers were eligible for employ e r- s p o n s o r e dhealth insurance in comparison to 58 per cent ofworkers in standard arrangements.23 For employersponsored pension plans, 8 per cent of temporaryhelp agency wo r kers and 31 per cent of on callwo r kers were eligible in comparison to 50 per centof workers in standard arrangements.

Self-employment

S e l f - e m p l oyment is the third major category ofnon-standard work in developed countries. Ove r a l lin 1997, for all of the OECD countries, self-employment represented about 12 per cent of non-agricultural employment, up from about 10 percent in 1979. 24 At the high end of these countrieswere Greece, Korea and Tu r key with a share ofaround one-quarter of non-agricultural employ-ment. At the low end were Austria, Denmark, France,G e r m a ny, Japan, New Zealand, Norway and theUnited States with less than 10 per cent of the non-agricultural labour force in self-employment.

S e l f - e m p l oyment is non-standard because itdoes not entail a wage relationship and, therefore,does not entail access to wage or payroll-basedbenefits. Put another way, the self-employed haveto depend upon their own profits and are person-ally responsible for their own welfare and that oftheir enterprise. T h ey do not have access toe m p l oy e r-sponsored benefits and often havelimited access to statutory or privatised systems ofsocial protection.

But the self-employed are not a homogenousgroup. Statistics on self-employment distinguishthree main sub-categories: a) self-employed withe m p l oyees (i.e. employ e r s ) ; b) self-employ e dwithout employees (i.e., own account operators,including independent contractors or free lancers);and c) unpaid family wo r kers. Unpaid fa m i l ywo r kers are usually excluded from compilations ofthe self-employed because they are consideredentrepreneurial assistants and not entrepreneurs.This would underestimate the real level ofwo m e n ’s entrepreneurship because some wo m e nc l a s s i fied as unpaid family wo r kers would be betterconsidered as partners with the self-employ e dp e r s o n .2 5 Within self-employment, ow n - a c c o u n twork is most directly relevant to the informal



economy because it is often characterized by weakattachments to formal structures, lower wages andpoorer working conditions. In OECD countries,f o r example, compared to both self-employ e de m p l oyers and regular employees, ow n - a c c o u n tworkers tend to report poorer working conditions,longer and irregular working hours, less access totraining, less autonomy and more job insecurity.

Recent changes in labour market structuresh ave made it harder to draw a clear distinctionbetween wage work and self-employment: this isbecause it is often difficult to verify whether aworker is dependent or independent. For example,s e l f - e m p l oyed wo r kers may receive work ordersfrom a formal business – their business customer– but perform the work at their customer’s bu s i n e s ssite and under the supervision of their bu s i n e s scustomer. In such cases, they are similar to depen-dent wage wo r kers. Moreover a number of gove r n-ments have been concerned about the rise of“ fa l s e ” self-employment in order to reduce taxliabilities. In such cases, the customer firm tries toavoid payroll taxes that cover unemployment insur-ance, wo r ker compensation insurance, pension,and (except in the United States) health insurance.The firms bypass all of these obligations bydeclaring that their wo r kers are self-employ e d . 2 6

Both self-employment as a whole and ow n -account self-employment as a share of total self-e m p l oyment have been growing. The proportion ofown-account self-employment within total self-e m p l oyment increased in six out of ten OECD

countries (where data were available) between1990 and 19972 7. Further, in most OECD countries,more self-employed wo r kers are own-account thanemployers. For example, in 1997, nine out of tens e l f - e m p l oyed jobs in Belgium and seven out of tenin Greece were own-account.

Women comprised one out of every three self-employed workers for OECD countries overall in1997 and this proportion is growing. From 1990 to1997 annual average growth rates of female self-e m p l oyment surpassed those of men in 10 out of18 countries where data are available. To someextent, this mirrors the increase in the proportionof women in total employment as a whole ;h ow ever the growth rate of women in self-employ-ment outpaced that of women in total employmentin the 1980s, then stabilized in the 1990s. 28

Benefits

S e l f - e m p l oyed wo r kers are – by definition –responsible for their own contributions to nationalretirement plans (in countries with unive r s a lsystems where they are eligible to enrol), healthinsurance schemes, and other protection. In somecountries enrolment and payment of a premiumo u t of one’s resources is mandated by law. Fo rexample, most countries have a tax-financed oldage pension system with mandatory contributions.S e l f - e m p l oyed wo r kers are usually ineligible forg overnment-run unemployment insurance, beingconsidered responsible for generating their ow njobs.

NOTES1 The following technical considerations apply to the tables that are in the section of this chapter on developing countries: while there is noUnited Nations standard for defining a country or area as developed or developing, these terms are applied for statistical and analytical purposesand do not express a judgment about the stage of development that a particular country or area has reached. The basic grouping of countries isby continental region. The listing of countries for each of the regions and sub-regions is in Annex 1. Regional and sub-regional averages arebased on unweighted data for the countries and areas for which data are available. If country data were weighted by population, regional andsub-regional averages would mainly reflect the situation in one or two large countries. 2 The caveats in using self-employment as an indicator of employment in informal enterprises are: 1) data on self-employment also includessome employers in medium and large enterprises who are in the formal sector and high-end professionals; however, their numbers would besmall, particularly in developing countries and 2) the informal sector includes employees of informal enterprises who are not reflected in measuresof self-employment.3 In addition to agriculture which is not considered here, the other two main economic sectors are industry and services. Industry comprisesmining and quarrying, manufacturing, electricity, gas and water supply and construction. Services is disaggregated here into two categories: a)wholesale and retail trade and b) hotels and restaurants ; transport, storage and communications; financial, real estate and bu s i n e s s ; public admin-istration, social security and defense; education; health and social work; and other community, social and personal services. This is based onUnited Nations International Standard Industrial Classification of all Economic Activities, Revision 3 (United Nations publication, Sales No.E.90.XVII.II).4 However a trial estimate was prepared for Mexico, as reported in Chapter 3. 5 Jacques Charmes, “Women working in the informal sector in A f r i c a : new methods and new data” March 1998 available on the W I E G Owebsite (www.wiego.org).6 Edeltraund Hoffmann and Ulrich Wa lwei, “The Change in work arrangements in Denmark and Germany : Erosion or renaissance of standards?” Paper for Upjohn Institute’s Conference on Non-Standard Work Arrangements in Japan, Europe and the United States, (Michigan,August 2000).7 U.S. Bureau of Labor Statistics, Contingent and Alternative Employment Arrangements, February 2001. (Washington, D.C, U.S. Departmentof Labor, 2001).

8 Organization for Economic Co-operation and Development (OECD), OECD Employment Outlook. (Paris, OECD, 1999)9 Ibid.1 0 European Council Directive EU 91712175N (Council of Ministers 1997). As described in Francois Michon, “Temporary agency work inEurope, ” European Foundation for the Improvement of Living and Working Conditions, (Dublin 1999).1 1 I b i d . and Cranfield University School of Management, 1997. Working Time and Contract Flexibility in the EU: Main Report. Bedford, England: Cranfield University School of Management. Cited in OECD 1999.]1 2 O rganization for Economic Co-operation and Development (OECD). OECD Employment Outlook. ( Paris, OECD,1997); Doudeijns, M. 1998.“Are Benefits a Disincentive to Work Part-Time?”, in O’Reilly, J. and Fagan, C. (eds.), Part-Time Prospect. Routledge, London and NewYork,pp. 116-136. Cited in OECD 1999.]13 Brenda Lipsett and Mark Reesor, “Flexible work arrangements: evidence from the 1991 and 1995 Survey of Work Arrangements”, HumanResources Development Canada Research Paper R-97 (Quebec, Canada, 1997). 14 Ken Hudson, “No shortage of ‘nonstandard jobs’nearly 30% of workers employed in part-time, temping, and other alternative arrangements”, Economic Policy Institute Briefing Paper (Washington, D.C., 1999). 15 Susan Houseman and Machiko Osawa, “The growth of nonstandard employment in Japan and the United States: a comparison of causes andconsequences”, Revision of paper presented at the Upjohn Institute’s Conference on Nonstandard Work Arrangements in Japan, Europe and theUnited States, Michigan, August 2000. 16 Hoffman and Walwei, op. cit.1 7 Colette Fa gan and Kevin Ward, “R egulatory conve rg e n c e? nonstandard work in the UK and the Netherlands.” Paper for the Upjohn Institute’sConference on Nonstandard Work Arrangements in Japan, Europe and the United States, Michigan, August 2000.18 Ibid.1 9 Houseman, Susan, and Machiko Osawa, ibid. Ta bulations by Houseman and Osawa from February 1999 Supplement to the Current PopulationSurvey. [Agency temporary workers accounted for 2.5 percent of all employed people.]20 Houseman and Osawa, op. cit.21 European Council Directive EU31999L0070 (Council of Ministers 1999) as described in Francois Michon, op. cit.22 Lipsett and Reesor, op. cit.2 3 Mehta, Chirag and Nik Theodore. 2001. “The Temporary Staffing Industry and U.S. Labor Marke t s : Implications for the Unemploy m e n tInsurance System”, paper prepared for America’s Workforce Network Research Conference, June 25-26, 2001, Washington, D.C.24 Organization for Economic Co-operation and Development (OECD). OECD Employment Outlook. (Paris: OECD, 2000). 25 Ibid. 26 Ibid.27 Ibid.28 Ibid.


Illustrative Case Studies 33

As discussed in Chapter 2, only two or three coun-tries have collected data that provide a basis formeasuring the conceptual whole of informale m p l oy m e n t : that is, both inside and outsideinformal enterprises. The first part of this chapterpresents three country case studies of recent eff o r t sto generate national estimates of the informale c o n o m y, including agriculture. The second part ofthis chapter reviews available statistics and quali-t a t ive studies on two important sub-groups ofwo r kers in the global informal economy: home-based workers and street vendors. The case studies of India, Mexico and South A f r i c ap r ovide an ove r v i ew of the size and compositionof the informal economy. In each of these coun-tries, informed data users requested and wo r ke dwith national statistical agencies to deve l o ps u r veys for the improved measurement of theinformal economy. National experts have subse-quently analyzed the data and prepared thef o l l owing reports. The joint efforts of producersand users to produce improved national data, asreported in these studies, is a critical building blocknot only for estimating the size and nature of theinformal economy worldwide but also for pointingto measurement issues and strategies. T h e s estudies show that the informal economy is a majorcomponent of total employment, ranging fromabout 35 percent of total employment in SouthAfrica to more than 60 per cent in Mexico andmore than 90 per cent in India. They also show theimportance of improved methods to measure notonly employment in the informal sector but also ininformal jobs outside the informal sector and,specifically, in agriculture.Home-based wo r kers and street vendors are two ofthe largest sub-groups of the informal workforce.Taken together, they comprise an estimated 10 to25 percent of the non-agricultural workforce ind eveloping countries and over 5 percent of the totalworkforce in developed countries. In most rega r d s ,these two sub-groups could not be more different.Street vendors are the most visible and most orga-nized group of the informal workforce, whereashome-based workers are the least visible and least

o rganized group. Yet street vendors and home-based wo r kers have several things in common.Both groups are global phenomena with modern aswell as traditional guises; both lack social protec-tion and employment benefits; both are describedin terms of the location of work yet encompassdifferent types of employment relations; and bothremain under-enumerated or poorly classified inlabour force statistics. Signifi c a n t l y, women aredominant in both groups: in developing countries,women represent 30 to 90 percent of street ve n d o r s( except in societies that restrict wo m e n ’s mobility)and 35 to 80 percent of home-based workers. Thissection of the booklet presents brief discussions ofthese two important sub-groups of the informalworkforce, including a summary of available dataon their respective size and on the gender dimen-sions of the phenomena.

Country Case Studies

India

India has enjoyed a growth rate in gross domesticproduct of about 6.5 per cent per annum over thelast decade. The impact of this consistent growthhas been reflected in a reduction in poverty sincethe mid-1980’s. However, a recent Task Force onEmployment Opportunities set up by the PlanningCommission in India in 2001 pointed out that eve na continuation of the GDP growth at this rate is notl i kely to bring about significant improvement inthe employment situation. 1 This is because muchof the growth has been capital- and information-intensive rather than labour-intensive. Also, someof the growth has been associated with informal,rather than formal, wage jobs. For example, therehas been limited growth in formal employment inthe manufacturing sector. Large proportions of thework force are in the informal economy in India,working as self-employed or in informal jobs.Without an appropriate policy environment, it isd i fficult for the benefits of economic growth toreach these categories of workers.

3. ILLUSTRATIVE CASE STUDIES


Size and Composition of Employment in theInformal Economy

The informal workforce in India is an estimated370 million wo r kers, nearly 93 per cent of the totalworkforce (table 3.1). The informal workforce iscomprised of three main segments. First, informale m p l oyment in agriculture – comprised of the self-employed in small-scale farm units and of agricul-tural labour – which continues to be important in

India, represents 60 per cent of total employment.Second, employment in informal enterprises/sectoroutside of agriculture represents another 28 percent of total informal employment. Third, informale m p l oyment outside informal enterprises andoutside of agriculture is an estimated 6 per cent ofthe total informal workforce (this is the ResidualCategory in table 3.1). Among the total non-agri-cultural workforce, 133 million workers or 83 percent are in the informal economy.

Table 3.1 Composition of employment in the informal economy in India, 2000

2000

Total Women Men

Total employment 1 397,720,000 123,270,000 274,450,000Total non-agricultural employment 159,897,000 31,061,000 128,836,000Total agricultural employment 238,197,000 92,308,000 145,889,000

Total informal employment2

Number of persons 369,755,000 118,220,000 251,533,000Percentage of total employment 93 96 92

Non-agricultural informal employment3

Number of persons 133,355,000 26,601,000 106,751,000Percentage of total employment 34 22 39Percentage of non-agricultural employment 83 86 83Percentage of total informal employment 36 23 42

Employment in informal enterprisesNumber of persons 110,034,000 22,200,000 87,834,000Percentage of total employment 28 18 32Percentage of non-agricultural employment 69 71 68Percentage of total informal employment 30 19 35Percentage of non-agricultural informal employment 83 83 82

“Residual”4

Number of persons 23,321,000 4,401,000 18,917,000Percentage of total employment 6 4 7Percentage of non-agricultural employment 15 14 15Percentage of total informal employment 6 4 8Percentage of non-agricultural informal employment 17 17 18

Agricultural informal employmentNumber of persons 236,779,000 91,723,000 145,056,000Percentage of total employment 60 74 53Percentage of agricultural employment 99 99 99Percentage of total informal employment 64 78 58Source : Calculations prepared by Jeemol Unni from national census and survey data1 Refers to population age 5 and above2 Total informal employment includes both agricultural and non-agricultural informal employment.3 Non-agricultural informal employment consists of employment in informal enterprises plus the “residual”.4 The «residual» equals total non-agricultural informal employment minus employment in the informal sector


The 23 million wo r kers in informal employ-ment outside informal enterprises and outside ofagriculture include outwo r kers or homewo r ke r swho are sub-contracted by formal sector units anddomestic workers engaged by households, such asmaids, gardeners and security staff. This categoryis determined by a residual method which esti-mates workers in informal jobs who are excludedfrom the official definition of the informal sector(see box on estimating employment in thei n formal economy in India). It is likely that theresidual method also identifies some workers whowould fall under the informal sector definition butare incorrectly excluded – or missed – in informalsector statistics. For example, there is evidence thatcasual wo r kers for informal firms, particularly inthe construction industry, outworkers for informalfirms, and barbers, tailors, or other service wo r ke r swithout a fi xed place of work are not fully capturedin statistics on informal enterprises. With no fixedplace of work, these wo r kers are often missed indata collection efforts.

Women account for about one-third of thewo r kers – about 118 million – in the informaleconomy while men account for two-thirds – about252 million. Women informal wo r kers are concen-trated mainly in agriculture – so much so thatthree-quarters of all employed women are ininformal employment in agriculture. An additional

18 percent of all employed women work ininformal enterprises. Over one half (53 per cent)of all employed men are in informal agriculturale m p l oyment and 32 percent are informal enter-prises.

Informal employment accounts for virtually alle m p l oyment in agriculture and in trade (99 per centfor both industries) and the overwhelming share ofemployment in construction (94 per cent). It alsoaccounts for the majority of workers employed intransport and storage (79 per cent) and in socialand personal services as well as financial service(66 per cent in both industries). How eve r, thenumber of wo r kers in the financial service categ o r yis very small. When informal agriculture isincluded, the overwhelming share of wo r kers in theinformal economy are in rural areas (78 per cent).By contrast, among non-agricultural workers, theshare in rural and urban areas is almost eve n ;55 per cent are located in urban and 45 per cent inrural areas.

Outside of informal employment in agricul-ture, wo r kers in the informal economy are mostl i kely to be employed in manufacturing, trade, andhotels. Among non-agricultural informal workers,43 percent of women are employed in manufa c-turing and about 28 per cent in trade; while about28 per cent of men are employed in each of thegroups.

Estimating Employment in the Informal Economy in IndiaInformal employment is defined as the sum of employment in the informal sector (an estimate

based on the characteristics of the enterprise) and in informal jobs outside informal enterprises(estimated through a residual method). The residual is derived through a comparative analysis ofstatistical sources.

To estimate the size of total informal employment, the specific steps are as follows :1. Determine the size of the total workforce by applying economic participation rates obtained

f rom the Employment and Unemployment Surv e y, NSSO (2001a) to population census esti-mates interpolated for the reference year.

2. Determine the size of employment in the organised sector based on administrative data of theD i rectorate General of Employment and Training (DGE&T) collected through a network ofemployment exchanges in the country that refers to all establishments in the public sector andenterprises in the private sector with 10 or more workers. Information from public sector andprivate establishments hiring 25 or more workers is mandatory under the law; however, privateenterprises employing 10-24 workers file this information on a voluntary basis. This could leadto an underestimation of the organised sector work force.

3. Subtract employment in the organised sector from the total workforce to determine total employ-ment in the informal economyTo further estimate the size of employment in informal jobs, outside of informal enterprises:

4. D e t e rmine employment in the informal sector based on responses to a question in theEmployment and Unemployment Survey on the registration status of the enterprise in which theworker is engaged.

5. Subtract employment in the informal sector from total employment in the informal economy tod e t e rmine i n f o rmal employment outside informal enterprises – which is a proxy for totalemployment in informal jobs .

Contribution of the Informal Sector to NationalIncome

The National Accounts Statistics of India estimatethe contribution of the organised and unorganisedsectors to the Net Domestic Product (NDP) : thatis, the Gross Domestic Product minus depreciation.The organised or formal sector is based on thec o n t r i bution of gove r n m e n t / s e m i - g ove r n m e n to rga n i z a t i o n s ; enterprises registered under theIndian Factories Act, Bidi and Cigar Workers Act,C o - o p e r a t ive Societies Act, Provident Fund A c t1 9 6 6; and recognized educational institutions. T h eunorganised sector is comprised of all enterprisesoutside those listed above.

In 1997-98, the contribution of the unorga n i z e ds e c t o r, including agriculture, was 60 per cent oftotal NDP. Excluding agriculture, the unorganizedsector contributed 45 per cent of non-agriculturalN D P. Estimates of the contribution of women ininformal employment indicate that wo m e ninformal wo r kers contribute about 32 per cent tototal NDP when agriculture is included and only20 per cent to non-agricultural NDP. By industry,the share of informal women wo r kers is about27 per cent in manufacturing NDP and 34 per centin community, social and personal services NDP.

In spite of the significance of these contribu-tions, they are substantially less than the share ofthe informal economy in employment which,including informal agriculture, is 93 per cent oftotal employ m e n t ; and, excluding informal agri-culture, 83 per cent of non-agricultural employ-ment.

Using conventional measures of productivity,unorganised workers would be considered to havel ow productiv i t y. In manufacturing, for instance,the unorganised sector contributes only about37 percent of total NDP but represents about 86 oftotal employment. In some industries, how eve r, thed i fference between the unorganised sector’s contri-bution to NDP and its share of total employ m e n tis not as great. In trade, the unorganised sectorc o n t r i butes about 84 per cent of the NDP andrepresents 99 percent of total employment. Intransport, the unorganised sector accounts forabout 62 per cent of the NDP and 70 percent oftotal employment.

In India, official statistics do not collect dataon the incomes of workers. As a proxy for income,the value-added per wo r ker computed from data inthe Informal Sector Survey can be used. T h eaverage annual value-added per wo r ker in allinformal enterprises was US $505 in 1999-2000.The scale of operation of the enterprise had a great

e ffect on the capacity to earn and generateincomes. While wo r kers in own account enter-prises earned only $375 per year, those in informalestablishments with hired wo r kers earned $816.F u r t h e r, in urban areas wo r kers in both ow naccount operations and informal establishmentsobtained higher incomes per worker than those inrural areas. In rural areas the average annualincome of workers in all informal enterprises wasbelow the poverty norm. Workers in own accountenterprises earned very low incomes and were themost vulnerable while, on average, wo r kers ininformal establishments (employers and employ e e st a ken together) obtained incomes above thepoverty norm. In urban areas, the average incomeof wo r kers in all informal enterprises was abovethe urban poverty norm; however, the earnings ofown account wo r kers in urban areas generally werenot above the poverty norm.

Mexico

The 1990s were a period of great economic changein Mexico. Restructuring programmes early in the1990s resulted in more liberal financial policy andgreater privatisation of the economy. In 1994 theNorth American Free Trade Agreement (NAFTA)was signed. A financial crisis occurred at the endof 1994 and GDP fell 6.2 per cent for the year –thelargest drop during the century. A quick economicr e c overy followed with sustained economic grow t hfor the latter half of the decade. At the end of thedecade the informal economy continued to beimportant to both employment and the GDP ofMexico.

Size and Composition of Employment in theInformal Economy

In 1998 informal employment accounted for 64 percent of total employment in Mexico (table 3.2).The non-agricultural component represented45 per cent of total informal employment and wasevenly divided between employment in theinformal sector/enterprises (23 per cent of totale m p l oyment) and informal employment outsideinformal enterprises (22 per cent of total employ-ment). Informal employment in agriculture wa s19 per cent of total employment and 94 per cent ofagricultural employment (table 3.3) (see box onestimating informal employment and thec o n t r i bution of the informal economy to GDPfor Mexico).

In terms of absolute numbers, twice as manymen, as women, worked in informal employmentin the year 2000 – 16 million men compared to



Table 3.2 Informal Economy in Mexico, 1998: Employment and Contribution to GDP1

Share Share Share of Non-Employment of Total GDP2

of GDP agricultural Employment GDP

Informal Enterprises 8,858,481 23.0 445,441,361 13.0 13.4Informal Employment Outside

of Informal Enterprises 8,480,252 22.0 494,754,798 14.0 14.8Non-Agricultural Informal

Employment 17,338,733 45.0 940,196,159 27.0 28.2Agricultural Informal Employment 7,334,412 19.0 172,132,923 5.0 –.–Total Informal Employment 24,673,145 64.0 1,112,329,082 32.0 –.–S o u rc e: Calculations pre p a red by Rodrigo Negrete, Coord i n a t o r, Office of Technical Direction, Instituto Nacional de Estadística,Geografía e Informática (INEGI).1 All calculations were based on main activity; secondary jobs were not considered here.2 Numbers refers to current pesos in thousands.

Estimating informal employment and contribution of informal economy to GDPfor Mexico

The data summarised in this study for i n f o rmal sector/enterprises a re based on NationalS u rveys of Micro-Business (ENAMIN) which covered a sample of all small-scale employers andown account workers previously identified in the National Survey of Urban Employment. The micro -business surveys were conducted on a bi-annual basis during the 1990’s until 1998. In order tohave national data and to have annual data, links were established to the National EmploymentS u rveys which have been carried out yearly since 1995. The criteria applied to data from theENAMIN to determine informal sector economic units were as follows: a) all “own account” enterprises were included;c) units headed by professionals were included only if they had no accounting records;d) m i c ro-enterprises of less than 15 persons with no conventional accounting pro c e d u res were

included.Non-agricultural informal employment outside informal enterprises: data were based on ques-

tions in the National Employment Surveys on social security protection.The portion that corresponds to the informal sector employment was subtracted from all employ-

ment unprotected by social security. Agricultural informal employment : The National Employment Surveys also collected data on

agricultural employment that does not have social protection. This category includes such workersas well as those working on subsistence agricultural units and micro-farms.

Trial calculations were prepared on the size of the GDP corresponding to the total informaleconomy . While informal sector GDP is a standard concept, the value added contributed by thetotal informal economy is a promising new illustrative indicator.

The starting point for the calculation of the total contribution of the informal economy (TIE) isthe household GDP in the national accounts minus the Imputed Net Rent of Owner- o c c u p i e dDwellings by institutional sector, both concepts published by the Instituto Nacional de Estadística,Geografía e Informática (INEGI). Once this subtraction was made, the residual is added to theshare that Non-Agricultural Informal GDP Outside the Household Sector (H.S.) represents in termsof employment in the SNA’s total estimate of salaries and wages:

1) TIE = (SNA’s Household´s GDP* – INRD) + SNA’s Salary and Wages (Informal Outside the H.S./Wage Workers)

Where: TIE: Total Informal Economy ContributionINRD: Imputed Net Rent of Owner’s Occupied Dwellings* Published data: Includes Informal Sector’s GDP.

To obtain the share that corresponds to Non-Agricultural Informal GDP outside the Inform a lSector it was necessary to make explicit the share of Informal Sector GDP and the correspondingshare of Agricultural GDP:

2) Non-Agricultural Informal GDP Outside the Informal Sector = TIE – Inf. Sector’sGDP + Agriculture GDP (Ag. Inf. Employment / Total Ag. Employment)]

Thus: 3) Non Agricultural Informal GDP= Non Agricultural Informal GDP Outside the Informal Sector + Informal Sector GDP


7 . 7 million women. Because fewer women thanmen are in the labour force, the gender gap isn a r r ower in terms of the relative shares of informale m p l oyment in total employ m e n t : 58 per cent ofwomen wo r kers are in informal employment incontrast to 64 per cent of men wo r kers. For bothwomen and men, informal employment is about55 per cent of non-agricultural employment. Also,for both women and men, informal employ m e n toutside of agriculture is almost evenly div i d e dbetween employment in informal enterprises andinformal jobs outside informal enterprises.

There is some evidence that a significant shareof informal employment outside informal enter-prises consists of a) disguised employee statuswhereby paid wo r kers work off site on wo r ksupplied by firms/employers under sub-contracts;or b) temporary worker status whereby employeesagree to renew their contracts with their employ e r severy three to six months: thus becoming “peren-nial temporary wo r ke r s”. Under both such arrange-ments, the employee is not entitled to a full rangeof labour benefits because a) the employer candisavow his/her responsibility to the employee; orb) the employee cannot accumulate a long enoughperiod in the job to qualify for benefits.

Many young workers without families acceptthese arrangements on the understanding that theywill be compensated in cash earnings for at leastpart of what the employ e r, under a standardemployment contract, would have to contribute tothe Social Security System. The employers preferthis arrangement because they have no othero b l i gations to their “e m p l oy e e s” beyond wa g epayments in cash.

Other examples of informal jobs outsideinformal enterprises include the traditional cate-gory of domestic workers; Mexicans living on theborder and working with no labour benefits in theUnited States, mainly in the service sector; and, arelatively new category, sales agents who work oncommission for legally constituted enterprises butare not listed in the firm’s records.

In Mexico, informal employment is a moreimportant source of employment in rural than inurbanised areas. In 2000, informal employ m e n trepresented about three quarters (77 per cent) ofe m p l oyment in areas with less than 100,000 inhab-itants and a little less than one half (46 per cent) inareas with more than 100,000 inhabitants. T h eproportions are roughly similar for women andfor men.

Informal employment plays a major role ineach of the main industry groups. It is the ove r-whelming majority of those employed in three

major industry groups: agriculture (94 per cent),construction and trade (both 73 per cent). In addi-tion, it is 63 per cent of transportation, 48 per centof communal, personal and other services, and2 2 per cent of financial services, insurance andreal estate.

For women, informal employment dominatedtheir participation in agriculture (94 per cent) andtrade (80 per cent). For men, informal employ m e n talso predominated in agriculture (94 per cent) andtrade 67 (per cent) but it was also significant inconstruction (79 per cent), communication andtransport (67 per cent) and communal, personaland other services (52 per cent). However, only alittle more than a third of men working in manu-facturing were in informal employment but abouthalf of women in manufacturing jobs were ininformal employment.

Contribution to GDP

With respect to GDP, trial estimates were preparedwhich show that the i n formal economy ( i n c l u d i n ginformal agriculture) contributes 32 per cent ofMexico’s GDP (see box on estimating informalemployment and contribution of info r m a leconomy to GDP for Mexico). The non-agricul-tural informal economy accounts for 27 per cent ofthe total GDP, with 13 per cent attributed to theinformal sector and 14 per cent to non-agriculturalinformal employment not linked to the informalsector.

The contribution of the informal sector to GDPis the more standard calculation in nationalaccounts. In Mexico, the 13 per cent contributionof the informal sector to GDP is almost three timesthe contribution of the agricultural sector to thee c o n o m y. More than half of the value added by theinformal sector is from trade activities (51 percent), followed by services (25 per cent) and manu-facturing (14 per cent).

Wo m e n ’s share of informal sector GDP in1999 was 33 per cent. By industry, the share ofwomen in trade GDP was almost equal to the shareof men (47 per cent for women and 53 per cent formen). Women also contributed a large share inmanufacturing GDP (38 per cent in comparison to62 per cent for men).

Per capita income from the Mexican InformalSector (informal sector GDP divided by employ-ment in informal enterprises) is well above theM exican minimum wage and the gap widenedbetween 1995 and 1999. Per capita income wa s4 . 5 times the minimum wage in 1995 and increasedto 5.2 times in 1999. For men, the ratio wentf r o m 4.9 in 1995 to 5.8 in 1999, and for wo m e n

from 3.8 to 4.2. These comparisons provide aninteresting reference point, however it should alsobe noted that the minimum wage lost itspurchasing power through the 1990s.

South Africa

The development of the informal economy inSouth Africa has been closely tied to the politics ofapartheid. Under apartheid, most informal sellingin urban centres and even would-be formal black-owned businesses were defined as illegal. Womenwere especially affected by the spatial laws ofapartheid. For those not born in ‘white’ u r b a nareas, residence was dependent on their relation-ship with a man – a husband or fa t h e r. Fo r m a lsector opportunities were also difficult to obtain.All of these conditions encouraged large numbersof African women to work as domestic wo r ke r s .

During apartheid, ‘informal’, ‘black’ and ‘illegal’were often treated as synonyms in describingeconomic activities that were formally discouragedby the apartheid government, but arose as aneconomic response by black people to apartheidpolicies. Since the formal ending of apartheid, theinformal economy has grown. Restrictions onb l a c k - owned businesses have been loosened andthe growth of small black-owned businesses hasbeen encouraged. How eve r, the concept of the“ informal sector ” is still often confused andconflated with illegal activities.

Size and Composition of Employment in the Informal Economy

E m p l oyment in the informal economy in SouthAfrica was estimated to be 34 per cent of totale m p l oyment in the year 2000 (table 3.4). T h i sincludes employment in the informal sector –


Table 3.4 Composition of employment in the informal economy in South Africa, 2000

2000

Total Women Men

Total employment 11,946,000 5,434,000 6,511,000Total non-agricultural employment 10,110,000 4,581,000 5,528,000Total agricultural employment 1,836,000 853,000 983,000

Total informal employment1

Number of persons 4,063,000 2,449,000 1,613,000Percentage of total employment 34 45 25

Employment in informal enterprises Number of persons 3,059,000 1,486,000 1,572,000Percentage of total employment 26 27 24Percentage of total informal employment 75 61 97

Non-agricultural employment in informal enterprisesNumber of persons 1,977,000 888,000 1,088,000Percentage of total employment 17 16 17Percentage of non-agricultural employment 20 19 20Percentage of total informal employment 49 36 67

Agricultural employment in informal enterprises Number of persons 1,082,000 598,000 484,000Percentage of total employment 9 11 7Percentage of agricultural employment 59 70 49Percentage of total informal employment 27 24 30

Employment in paid domestic workNumber of persons 1,004,000 963,000 41,000Percentage of total employment 8 18 1Percentage of total informal employment 25 39 3S o u rc e: P re p a red by Debbie Budlender, Peter Buwembo and Nozipho Shabalala, from the 2000 Labour Force Surv e y, StatisticsSouth Africa.1 Informal employment consists of employment in informal enterprises and paid domestic work.

2 6 per cent of total employment – and employ m e n tin paid domestic work – 8 per cent of total employ-ment. It should be noted that South Africa includesagriculture in these measures. How eve r, indiv i d-uals working in informal jobs for formal firms arenot included in these estimates. For this reason, 34per cent is a conserva t ive estimate of the size of theinformal economy in South Africa. Additional datacollected in the labour force survey indicate thatmany workers in the formal sector are in jobs withinformal working conditions, specifically jobs

a) without a written contract; b) under temporary,casual or seasonal terms of employ m e n t ; andc) with no entitlement to paid leave. However, inthe view of the South Africa analysts, the conceptof informal employment is best seen as acontinuum (see box on the formal and informalsectors as parts of a continuum). For this reasonit was not possible to combine the data from thed i fferent survey questions to arrive at a singleestimate of the total informal economy for thisreport.


The formal and informal sectors as parts of a continuumThe Labour Force Survey 2000 determined whether a worker was in the formal or informal

sector by asking whether the business where the individual worked was in the formal or informals e c t o r. Respondents were also asked about characteristics of the organisation, business, enterpriseor branch where the individual worked and, in the case of employees, about the characteristics ofthe job.

With respect to characteristics of the enterprise where the individual worked, questions wereasked about: a) size (in terms of number of workers) ; b) whether the enterprise is a re g i s t e re dcompany; c) whether the enterprise deducts Unemployment Insurance Act (UIF) contributions; andd) location of the enterprise, specifically in the owner’s home or farm, in someone else’s home,inside a formal business, at a service outlet, at a market, on a footpath, street, etc. The re s u l t sshowed there was a clear relationship but not an exact correlation between the characteristics andthe formality of the enterprise. For example, 80 per cent of informal enterprises have no charac-teristics of formal enterprises. Conversely, only 1 per cent of informal enterprises have all four char-acteristics of formal enterprises, compared to 42 per cent of formal sector enterprises. The patternsuggests there are degrees of formality and informality along a continuum rather than mutuallydistinct sectors.

The Labour Force Survey also examined the formality of employment relationships amongworkers classified as employees by asking about a) the presence of a written contract; b) the dura-tion of employment; and c) entitlement to paid leave. Here again – whether analysing the alter-natives individually or as combined conditions of a job – there was a clear relationship, but notexact correlation, between the characteristics and whether workers had defined themselves asworking in formal or informal enterprises. For example, 9 per cent of employees in the formal sectorworked under all three of the conditions of informal work in comparison to 36 per cent of domesticworkers and 50 per cent of those in the informal sector. Moreover, 55 per cent of employees inthe formal sector had none of the working conditions that are associated with informality in compar-ison to 5 per cent of domestic workers and 10 per cent of informal sector employees.

T h ree establishment surveys conducted by the World Bank in the Greater Johannesburgmetropolitan area also show the links between the formal and informal sectors.

The study of large firms found a positive relationship between firm size and the use of tempo-rary labour and outsourcing. Over three-quarters of firms using temporary labour said they did soin order to expand the workforce without hiring permanent workers, what the report termed “flex-ibility” in adapting to increased workloads. A second study looked at small but formal firms regis-t e red for payment of Value Added Tax (VAT). Again citing the need for flexibility, a third of themicro firms (1 to 5 employees), 50 per cent of very small firms (6 to 20 employees) and 60 percent of small firms (21 to 49 employees) hired temporary labour. A third study covered informalfirms—firms which were not registered for VAT and not subject to other formal regulation or taxa-tion. Retail and hawking accounted for about one-quarter of the firms. The study found that 80 percent of the informal firms depended on formal firms for inputs, 36 per cent sold directly to formalfirms, and over 50 per cent competed with them. Sub-contracted work was not as important forthese informal firms; only 14 per cent had received sub-contracted work.Sources: V. Chandra, L. Moorty, B. Rajaratnam and K. Schaefer, “Constraints to growth and employment in South Africa,” Report no.1: Statistics from the Large Manufacturing Firm Survey, The World Bank Southern Africa Department: Informaldiscusson papers on aspects of the economy of South Africa, Discussion paper 14 ( June 2001) and Ibid., Report No 2 :Evidence from the Small, Medium and Micro Enterprise Survey. Discussion paper 15 (June 2001). Monique Cohen, with Mihir Bhatt and Pat Horn, “Women Street Vendors: The Road to Recognition, Seeds, No. 20, pp.1-23, (New York, Population Council, 2000).

More women are employed in the informaleconomy than men – 2.4 million women incomparison to 1.6 million men (table 3.4).M o r e ove r, nearly one-half of all employed wo m e nwork in the informal economy, but only one-quarter of employed men. The number of womenand men working in the informal sector is roughlys i m i l a r. More women than men work in theagricultural informal sector and substantially morework in paid domestic work. Domestic service is39 per cent of wo m e n ’s total informal employ m e n tbut only 3 per cent of men’s (see box on domesticworkers).

Informal employment is more important innon-urban (in South Africa this includes the trulyrural and the urban periphery) than in urban areas.O ver half of employment in non-urban areas( 5 3 per cent) is informal but only one quarter(27 per cent) in urban areas.

There are also significant differences ine m p l oyment across the provinces, reflecting,among other factors, the country’s apartheidhistory and poverty levels. In the poorest prov i n c e swhich mainly comprise the previous “homeland”areas, Eastern Cape and Northern Prov i n c e(recently renamed Limpopo), formal employmentaccounts for under half of total employ m e n t .C o nve r s e l y, in Western Cape and Gauteng, the twowealthiest provinces, the formal sector accountsfor about three-quarters of employment. EasternCape and Northern Province are also among thep r ovinces with the highest levels of unemploy m e n t(27 per cent and 28 per cent, respectively) whileWestern Cape and Gauteng have the lowest (15 and21 per cent). In provinces which do not consistl a rgely of former “h o m e l a n d” areas – We s t e r n

Cape, Northern Cape, Free State and Gauteng –there is very little difference in the formal/informaldistinction in urban and non-urban areas. It isprimarily in the provinces which consist largely offormer “h o m e l a n d” areas that the non-urban areash ave significantly more informal employment. Al a rge part of the informal employment in theseprovinces is comprised of subsistence agriculturalworkers.

Informal employment as currently measured inSouth Africa is concentrated in two economicsectors: namely, domestic services and in agricul-ture. Together they account for 52 per cent ofemployment in the informal economy, as definedby South Africa. Construction and trade togetheraccount for another significant proportion – about30 per cent – of informal employment. Wo m e neven more than men are likely to be employed ininformal agriculture and trade. Women represent60 per cent of informal wo r kers in trade and 55 percent or more of informal wo r kers in agriculture,m a n u facturing, and community and personalservices.

Data are also available on occupation. Fewclerks, professionals, technical personnel and oper-ators are employed in the informal sector. Bycontrast, the majority of farmers, gardeners andother skilled agricultural workers (81 per cent) arein the informal sector. In addition signifi c a n tportions of unskilled wo r kers (36 per cent), craftwo r kers (32 per cent) and service and salesworkers (27 per cent) are in the informal sector. Ineach of these occupational groups, women aremore likely than men to be employed in theinformal sector. Women account for roughly onehalf of wo r kers in the informal sector (ex c l u d i n g


Domestic workersAs noted before, domestic workers are a significant source of employment in South Africa,

especially for women (see table 3.4) . Domestic workers comprise 8 per cent of total employmentand 18 per cent of women’s total employment. The South Africa Minimum Standards Division ofthe Department of Labour carried out a study of domestic workers for the Employment ConditionsCommission, the body which advises the Minister of Labour on minimum wages and conditionsfor vulnerable workers. The department drew on all known existing work on domestic workers andalso organised some special investigations in compiling the report. The sources included employerq u e s t i o n n a i res, public workshops and hearings, a telephone survey of urban households, ques-tionnaires completed by domestic workers and information from Statistics South Africa.

In terms of age, domestic workers were found to be concentrated in 30 – 49 age group butnot confined to these ages. Generally, domestic workers have relatively low education levels whenc o m p a red to other workers. However there are significant numbers of women who have completeds e c o n d a ry education but are doing domestic work in the absence of other opportunities. Ones u rvey found that 37 per cent of workers said they were doing this work because they did not haveadequate skills for other occupations; 34 per cent said it was easier to find work as a domestic,and only 14 per cent reported that they did the work because they enjoyed it.

domestic work), but a higher proportion in severaloccupational groups : two-thirds of informalservice and sales workers, 62 per cent of informalsector clerks, 55 per cent of informal sectorunskilled wo r kers and 53 per cent of techniciansand associate professionals in the informal sector(including wo r kers in computing, nursing,m i d w i f e r y, traditional medicine and teaching aswell as farmers, gardeners and other skilled agri-cultural workers.

Sub-Group Case Studies

Home-Based Work

Working at Home

Home-based work appears to be on the rise aroundthe world. Some combination of the follow i n gfactors accounts for this apparent increase. T h efirst is that global competition increases pressureson firms to cut costs through more flexible wo r kcontracts or sub-contracting production. T h esecond is that information technology – particu-larly computers – allows and encourages manyclerical, technical, and professional wo r kers towork from their home rather than at another work-site. The third is that an increasing lack of formale m p l oyment opportunities – due variously to thelack of economic growth, to capital-intensivepatterns of economic growth, and/or to fa s t e rgrowth in the economically active population thanin formal employment – forces many wo r kers tot a ke up self-employed work, often at or fromthe home.

Working at home evo kes two contrastingimages: one more traditional and pessimistic, theother more modern and optimistic. The pessimisticimage is of low-paid and low-skilled manual workdone in cramped, dingy and unsafe surroundings,often involving child labour. The optimistic imageis of highly paid and skilled professionals, techni-cians, and managers conducting business “by fax,phone, e-mail, and other computer links from thecomfort of their well-appointed residences”.2

In reality, home-based work is more heteroge-neous than these two prevailing images suggest.Home-based wo r kers include the self-employ e dwho are engaged in family businesses or ow naccount operations as well as paid wo r ke r sworking under sub-contracting arrangements.Some observers also include those who do someof their paid work at home although they also havea place of work outside their home. Some home-based wo r kers are engaged in labour- i n t e n s ive

manual activities, while others work in capital- andi n f o r m a t i o n - i n t e n s ive clerical or professionalactivities. Their occupations range from rolling orpackaging incense sticks and ciga r e t t e s ; tostitching garments or shoe uppers ; to prov i d i n gl a u n d r y, child care, or shoe repair services ; toassembling electrical plugs or electronic compo-n e n t s; to entering, processing, or analysing data ;to providing professional and technical services toindividuals or businesses.

Despite the considerable diversity in the termsand conditions of work experienced by diff e r e n tgroups of home-based wo r kers, there is grow i n gevidence to suggest why we should be concernedabout home-based workers, especially about thoseengaged in low-end work. One reason for concernrelates to a common problem faced by home-basedworkers and other informal workers: namely, thefact that they do not have access to employment-based benefits or protection. Another relates to thefact of working at or from home: that is, home-based workers tend to remain isolated from otherwo r kers and, therefore, to be less well orga n i s e dand have less voice vis-à-vis employers or publicauthorities than other wo r kers. Some reasons forconcern apply to specific groups of home-basedwo r kers, notably homewo r kers. Homework isoften associated with low pay, especially amongh o m ewo r kers engaged in manual wo r k .Compounding their often low wages is the fact thath o m ewo r kers have to pay for many of the non-wage costs of production : notably, the ove r h e a dcosts of space, utilities, and equipment (see box ons e l f - e m p l o yed, homewo r kers, and employe e s ).

Another reason for concern is that women areove r-represented among home-based wo r ke r s ,especially among homewo r kers engaged in manualwork. Available evidence from around the wo r l dsuggests the following common patterns: womenare more likely than men to work mainly at home;women are more likely than men to work at homein manual activ i t i e s ; and among homewo r ke r swomen are far more likely than men to be enga g e din low-paid manual work. 3 The available evidencealso suggests that women homeworkers in manualjobs are among the lowest paid wo r kers in theworld. 4

For these and other reasons, labour activ i s t sincluding the Self-Employed Wo m e n ’s A s s o c i a t i o n( S E WA) and the international alliance of home-based wo r kers (called HomeNet), lobbied for aninternational convention on homework that wouldmandate that homewo r kers be recognised aswo r kers and receive the rights due all wo r kers. T h eILO Convention on Homework, 1996 (No. 177)


recognises homewo r kers as wo r kers who areentitled to just reward for their labour and sets astandard for their minimum pay and wo r k i n gconditions, including occupational health ands a f e t y. Among other recommendations, the ILOC o nvention on Homework, 1996 (No. 1 7 7 ) calls forimproved statistics on homework.

Composition of Home-based Work

The term “home-based worker” is used to refer tothe general category of wo r kers who carry outr e m u n e r a t ive work within their homes or in thesurrounding grounds. It does not refer to eitherunpaid housework or paid domestic work. Withinthe general category of home-based wo r kers, thereare two basic types of wo r ke r s: those who work ontheir own (the self-employed) and those who wo r kfor others (mainly as industrial outwo r kers). T h eterm “h o m ewo r ke r” is used to refer to a sub-seto f home-based wo r ke r s : namely, industrialo u t wo r kers who carry out paid work from theirhome. It is important to distinguish, both concep-tually and statistically, between the following twocategories (and related terms):

■ Home-based wo r ke r s: all those who carry outm a r ket work at home or in adjacent grounds orpremises whether as self-employed or as paidworkers

■ H o m e wo r ke r s : those home-based wo r ke r swho carry out paid work for firms/businessesor their intermediaries, typically on a piece-ratebasis

This distinction reflects differences that havepolicy implications. The problems and constraintsfaced by self-employed home-based wo r kers andh o m ewo r kers are quite different, although bothtypically lack bargaining power and have toprovide their own social protection. Homeworkersare often forced by circumstances to work for lowwages without secure contracts or fringe benefitsand to cover some production costs (in particular,equipment, space, utility costs). Most self-e m p l oyed home-based wo r kers, except high-endprofessionals, face limited access to and/ orcompetition in relevant markets. To improve theirsituation homewo r kers need to strengthen theircapacity to bargain for regular work orders, higherpiece rates, and overdue back pay (a commonproblem faced by homewo r kers wo r l d w i d e ); whilehome-based self-employed need better accesst o financial markets and enhanced capacity tocompete in product markets. In effect, home-workers often face problems of exploitation whilethe self-employed often face problems of ex c l u-

s i o n. The strategies to address problems ofexploitation in labour markets – such as collectiveb a rgaining for higher wages – are different than thes t r a t egies to address problems of exclusion incapital and product markets – such as prov i d i n gaccess to financial, marketing, and bu s i n e s sservices.

Despite their numbers, and despite the grow i n ginterest in their situation, there are few good esti-mates of home-based wo r kers in general and few e rstill of homewo r kers in particular. This is due, inpart, to problems of enumerating work carried outin the home, especially by women. This is also dueto the fact that the “place of work” variable, usedto identify persons working at or near their home,is not included in many labour force and popula-tion surveys and, even when it is, the results haveoften not been tabulated in official statistical anal-y s e s . 5 In addition, to obtain the information neededto understand the nature and scope of the problemsthey face, home-based workers need to be classi-fied according to appropriate employment statuscategories and by industry or sector.

There are several specific problems related tomeasuring homewo r kers in particular. An impor-tant problem is the difficulty of determiningwhether a home-based wo r ker works under acontract or agreement for a specific fi r m; and, ifso, whether she/he is self-employed or a home-wo r ke r. This is because homewo r kers occupy ag r ey intermediate space between the fully inde-pendent self-employed and fully dependent paide m p l oyees (see box on self-employed, home-wo r kers, and employe e s ). Homewo r kers typicallyhave to absorb many production costs and associ-ated risks – including, buying or renting and main-taining equipment ; providing workspace andpaying for utility costs; and buying some inputs –often without help from their employers. Thus theirnet remuneration may be significantly less thanindicated by the piece-rates that they are paid. Forinstance, most garment homeworkers have to buyand maintain their own sewing machines, replaceneedles and oil, and pay for the electricity to runtheir machines and light their workspace. Mosth o m ewo r kers are also not directly supervised bythose who contract work to them, although they aresubject to delivery deadlines and to quality controlof the products or services they deliver. For thesereasons, they should be considered semi-depen-dent, not dependent, wage wo r kers. To identifyclearly which home-based wo r kers are home-wo r kers, labour force surveys and populationcensuses need to include sufficient and appropriatequestions regarding their contractual situation.


A second problem is the difficulty in identi-fying the specific firm for which the homeworkerworks and determining the characteristics of thatfirm. The current national and international stan-dards for measuring “status in employment” treatthe intermediary – the contractor – who suppliesr aw materials and receives the finished goodsa gainst payment for the work done as the“employer”. However, analytically, it is not clearwhich firm should be considered as the employerof the homeworker: the intermediary that directlyplaces work orders, the supplier that puts out workto the intermediary, the manufacturer thatoutsources goods from the supplier, or the retailerthat sells the goods? There is a parallel lega lp r o b l e m: namely, which unit in the chain shouldbe held accountable for the rights and benefits ofwo r kers down the chain? Many labour lawyers anda c t ivists argue that the lead firm that initially putout the work should be considered the equivalentof the employ e r. Operationally, the homewo r ke roften does not know which firm puts out the workor sells the finished goods.

A related problem is that a category of workerassociated with the sub-contracting of work – the

intermediary or sub-contractor – does not fi tc o nventional categories of employment status.Sub-contracting by a manufacturing or retail firmoften invo l ves one or more intermediaries andsometimes invo l ves a long, complex chain of inter-mediaries. These intermediaries – or sub-contrac-tors – typically receive work orders and raw mate-rials from firms or other intermediaries. They thenput out work to small production units or to home-wo r kers. Many such intermediaries are themselve sh o m e - b a s e d : that is, they store raw materials attheir home and allocate work orders from theirhomes. In addition to putting out work to others,some intermediaries operate small production unitst h e m s e l ve s: thereby, taking on the additional statusof self-employed outwo r ke r. Unlike the indepen-dent employer who hires others to work in his/herenterprise, the intermediary depends on a firm oranother intermediary for work orders and rawmaterials and usually sub-contracts, rather thanhires, wo r kers. Like an independent employ e r,h ow eve r, the intermediary assumes some economicrisk: notably, responsibility for storing raw mate-rials, overseeing the quality of production, andd e l ivering finished goods. For these reasons,


Self-Employed, Homeworkers, and EmployeesCurrent national and international statistical standards used to measure and classify “status of

employment” do not have enough categories to capture the range of employment arrangementsin today’s world. This is because all workers are thought to be either fully independent (self-employed) workers on the one hand (such as employers and own-account workers) or fully depen-dent workers on the other (i.e. as paid employees). However, many work situations do not fit neatlyinto these two basic categories. Rather they fall in a grey intermediate zone between being fullyindependent and being fully dependent. Consider the intermediate status of homeworkers as illus-trated in the table below:

Characteristics Self-Employed Homeworkers Employees

Contract sales contract employment contract employment contractRemuneration from sale of goods/ for work (typically for work

services piece rate) (time or piece rate)Contract with Self employer/intermediary Employer

Means of provided by self provided by self provided by employerProductionWorkplace provided by self provided by self provided by employerSupervision Autonomous indirect or no supervision direct supervision

To be able to identify and enumerate homeworkers, and other workers with interm e d i a t eemployment status, specific questions to probe the key variables in the left-hand column – notably,nature of contract, form of remuneration, place of work, and degree of supervision – need to bedesigned. The current national and international standards for classifying workers would need tobe re-examined to determine whether existing categories can be sub-divided to accommodate theseintermediate employment statuses or whether whole new categories that cut across existing oneswould need to be introduced.

intermediaries are better considered as semi-inde-pendent wo r kers, rather than fully independentemployers.

Depending on the number of intermediaries ina ny given sub-contracting chain, the links betweenthe homeworker and the lead firm for which theywork are often obscure. In long complex chains ofintermediaries, bargaining for higher wages iscomplicated by the distance between the home-wo r ker and the lead firm and the ambiguity ove rwho is responsible for providing higher wages ( s e ethe box on homewo r kers in global value chains).

Nature of Home-Based Work

H i s t o r i c a l l y, home-based work has alwa y sincluded skilled artisan production and entre-preneurial activities as well as low-skilled manualwork and survival activities. In recent decades, newforms of home-based work – often invo l v i n ghigher-skills, information technology, and higher-wages – have emerged. Currently, the va r i o u sforms of home-based work include :

■ M a n u facturing and A s s e m b l y: sewing, packing,routine assembly

■ Artisan Production: weaving, basket-making,embroidery, and carpet-making

■ Personal Services : laundry, beautician andbarber, shoe repair, dressmaking, lodging andcatering

■ Clerical Wo r k : typing, data processing, tele-m a r keting, bookkeeping, accounting, callcentre telephonists

■ Professional Wo r k : tax accounting, legal advis-ing, design consulting, computer programm-ing, writing, engineering, architectural, medical

Ty p i c a l l y, manufacturing and assembly wo r kinvolve low levels of skills, technology, and pay;routine and standardised tasks; and/or phy s i c a l l ydemanding effort. Professional work, on the otherend, tends to be varied, complex, and creative ;relying heavily on information technology;involving choice and discretion; and well paid, ifnot highly paid. 6 The other forms of home-basedwork fall somewhere in between.

Historically, most homework involved manualwork in labour- i n t e n s ive activ i t i e s : notably, int extiles, garment, and footwear manufa c t u r i n gindustries. Increasingly, homework also invo l ve sa c t ivities in the service and commercial sectors :notably clerical work in data processing, telecom-munication, and telemarke t i n g ; but also highlyskilled professional and technical consulting. As aresult, increasing numbers of homewo r kers, partic-ularly in developed countries, are in services andcommerce, not manufacturing (see box on recentshifts in homework in Chile). New forms ofhomework have emerged also in capital-intensivem a n u facturing industries. For example, recenti n n ovations in production technologies and tech-niques allows the automobile industry to sub-contract some part of the production process tohome-based enterprises and (even) homeworkers.A recent survey of homework in the U.K. foundthat specialised firms producing specific compo-nents for the automobile industry, including wiringsystems, seat cushions, and waterproof cove r s ,sub-contract to homeworkers. 7

Numbers Working at Home

Available statistics for developed countries suggestthat a significant and probably growing number ofwo r kers are home-based. Available statistics for


Homeworkers in Global Value ChainsIn global value chains in which the lead firm is a multinational firm based in an industrialised

country and the homeworkers are scattered across one or more countries, the links between thehomeworker and the lead firm for which she/he works become obscure. The following case illus-trates how complicated things can be in negotiating payment or wages due for completed work.

When a trade union organiser in Canada tried to help one immigrant Chinese garment workerget her back wages, she found that the garment worker did not know who she worked for as theman who dropped off raw materials and picked up finished garments drove an unmarked van.When the garment worker eventually found a tag with a brand label on it among her raw mate-rials, the trade union activist was able to trace the “label” from a retail firm in Canada to a manu-facturing firm in Hong Kong to an intermediary in Canada: in this case, the global value chainbegan and ended in Canada. When the local intermediary was asked to pay the back wages dueto the garment homeworker he replied: “Put me in jail, I cannot pay. The manufacturer in HongKong who sub-contracted production to me has not paid me in months.”Source: Stephanie Tang of UNITE, personal communication.


Recent Shifts in Homework in ChileThe Government of Chile has added special modules to measure homework to its national

labour force survey: once in 1997 in both rural and urban areas, and again in 2000 in urbanareas only. These survey findings show shifts in the composition of urban homework in the contextof a prevailing economic crisis and growing unemployment (8 per cent in 1997, 11.5 per cent in2000). There were shifts in homework in the urban areas over this period, including : drop in shareof homeworkers in total workforc e; drop in share of women among homeworkers; and, most signif-icantly, rise in share of men among homeworkers (from 20 to 27 per cent).

T h e re are several possible explanations for the increased share of men in homework, including:the informalisation of formal jobs, the flexibility of home-based work, and the emergence of new,higher status, forms of home-based work – involving information technology and the use ofcomputers – that might attract men to home-based work.

Between the two survey rounds, the share of services/ commerce in total homework rose from45 to 56 per cent and the share of services/commerce in male homework rose from 13 to 71 percent. These findings are consistent with the latter explanation: namely, that higher status forms ofhome-based work in information-intensive services might have attracted men to home-based work.

In 2000, among all urban homeworkers in Chile, women were over-represented in manufac-turing and retail trade (90 and 98 per cent, respectively) while men were over- re p resented in pro f e s-sional and technical jobs (82 and 66 per cent, respectively). Sources: Helia Henriquez, Verónica Riquelme, Thelma Gálvez and Teresita Selamé et al. “Home Work in Chile: Past andPresent Results of a National Survey”, SEED Working Paper No.8, (Geneva, International Labour Office, 2001) and HeliaHenriquez and Veronica Riquelme, El Trabajo a Domicilio en el 2000, unpublished report.

Table 3.5 Home-Based Workers in Fourteen Developing Countries: Number, Share ofNon-Agricultural Employment, Proportion Women

Total Home-Based WorkersCountries/Categories Number of Per cent of Non- Women as of Workers Home-based Workers Agricultural Workforce per cent of Total

Only Homeworkers CoveredChile (1997) 79,740 2 82Philippines (1993-5) 2,025,017 14 79Thailand (1999) 311,790 2 80

Only Self-Employed CoveredBrazil (1995) 2,700,000 5 79Costa Rica (1997) 48,565 5 45Morocco (1982) 128,237 4 79Peru (1993) 128,700 5 35

Both Categories CoveredBenin (1992) 595,544 66 74Guatemala (2000) 721,506 26 77India (1999-2000) 23,496,800 17 44Kenya (1999) 777,100 15 35Mexico (1995) 5,358,331 17 43Tunisia (1997) 211,336 11 38Venezuela (1997) 1,385,241 18 63

S o u rc e : Jacques Charmes, 2002 (personal compilation of the author on the basis of official labour force statistics and nationalaccounts). A subset of these data was published in ILO, On Measuring Place of Work (Geneva, 2002).

developing countries suggest that over 10 per centof non-farm workers in most countries and as highas 20-25 per cent in some countries are home-based.

D eveloping Countries : Working at home hasa lways been the reality of work for many people ind eveloping countries. How eve r, statistics on thisphenomenon remain very poor. Recent compila-tions of official statistics from the early to mid-1990s on home-based work from 14 deve l o p i n gcountries suggest that there is considerable varia-tion in the incidence of home-based wo r k( t a b l e 3 . 5 ). In Benin, which has made specialattempts to improve its official statistics in thisarea, the share of home-based work in non-agri-cultural employment was very high (66 per cent).In seven of the countries, home-based wo r ke r srepresented between 10 to 25 per cent of the non-agricultural wo r k f o r c e : Guatemala (26 per cent),India (16 per cent), Ke nya (15 per cent), Mex i c o(17 per cent), Philippines (14 per cent), Tunisia (11per cent ), and Venezuela (18 per cent). In one ofthese countries, the Philippines, only homewo r ke r swere counted, while in Guatemala, India, Kenya,M exico, Tunisia, and Venezuela attempts weremade to count all home-based workers, both thosewho are self-employed and homeworkers.

In the other six countries, the share of home-based wo r kers in non-agricultural employment wa squite small: Brazil (5 per cent), Chile (just under2 per cent), Costa Rica (5 per cent), Morocco (4 p e rcent), Peru (5 per cent), and Thailand (2 per cent).H ow eve r, for two of these countries – Thailand andChile – only homewo r kers were counted ; and inthe other four – Brazil, Costa Rica, Morocco, andPeru – only the self-employed were counted.

Perhaps the most striking fact is that the shareof women in home-based work was over 75 percent in seven of the countries, over 50 per cent inanother one country, and over 30 per cent in theremaining six countries. In the three countries thatonly counted (dependent) homeworkers, the shareof women was about 80 per cent. A recent nationalsample survey in India, specially designed to betterenumerate the informal economy, found a highincidence of home-based work overall as well asm a r ked urban-rural and male-female differences inthe incidence of home-based work (see box onworking at home in India).

D eveloped Countries: Working at home is alsothe reality of work for many people in developedcountries. In the mid-1990s, home-based workers– here defined as persons who work more than halfof their working hours at (or from) their home –


Working at Home in IndiaIn India in 1999-2000, the National Sample Survey Organization included a question in the

labour force survey on the “place of work.” The results show that about one quarter of all non-agricultural workers in informal enterprises work in their own dwelling. Within this overall picture,there are marked urban-rural and male-female differences. About 18 per cent of these workers arehome-based in rural areas, while only 6 per cent are home-based in urban areas. Well over halfof the female non-agricultural workforce in informal enterprises (57 per cent) works at home, whileless than one-fifth of the male non-agricultural workforce (18 per cent) works at home.

A recent sample survey of the workforce of Ahmedabad city in Western India brings out veryclearly the gender diff e rences in the location of work. The findings from that survey show thefollowing distribution of all male and female workers – both formal and informal – across differentwork sites:• 52 per cent of all women, compared to 8 per cent of all men, work at home• 18 per cent of all women, compared to 1 per cent of all men, work in others’ homes• 5 per cent of all women, compared to 23 per cent of all men, work on the streets• 3 per cent of all women, compared to 5 per cent of all men, work at construction sites • 22 per cent of all women, compared to 58 per cent of all men, work at factories, offices, or

workshopsThis study found that women operate nearly 70 per cent of the informal manufacturing activi-

ties, nearly 30 per cent of the informal service activities, and just under 15 per cent of the informaltrading activities and that the majority of all economic activities managed or operated by womena re home–based. For instance, virtually no women run small manufacturing units outside theirhomes and nearly three–quarters of women traders operate from their homes (rather than on thestreets). S o u rc e: Jeemol Unni., “ Size ,contribution and characteristics of informal employment in India” (2001). Paper will beavailable at http://www.ilo.org/public/English/employment/infeco/index.htm

represented between four to eleven per cent of thetotal workforce in eight out of twelve Europeancountries survey e d : Belgium, Denmark, France,G e r m a ny, Ireland, Italy, Luxe m b o u rg, andNetherlands. In the remaining four countries, theshare of home-based wo r kers in the total wo r k f o r c ewas as follow s : Greece (one per cent), Portuga l(four per cent), Spain (one per cent), and the U.K.(three per cent). For the twelve European countriest a ken as a whole, the share of home-based wo r ke r srepresented between four and five per cent of thetotal workforce. 8

In the United States, home-based work grewbetween 1980 and 1990, after falling significantlybetween 1960 and 1980. Factors in this grow t hinclude advances in information and communica-tion technology and the need to balance work andfamily by the growing numbers of two-career fa m i-lies. Based on 1991 national survey results one percent of all non-farm wo r kers wo r ked entirely athome. Two-thirds of these wo r kers were wo m e n .In contrast to on-site wo r kers, home-based wo r ke r swere more likely to be self-employed, to work non-standard hours, and to live in rural areas. Analysiss h owed that wo r kers who need or prefer flex i b l ework hours or to work at home – the disabled ;women, especially those with young children ; andthose living in rural areas with long commutes toon-site jobs – had greater representation amonghome-based wo r kers. How eve r, the associatedf l exibility came at a cost. The average hourlywages of home-based wo r kers of either sex wereb e l ow those of on-site wo r kers, even when onecontrols for employment status, hours worked, orurban/rural residence. 9 However, comparing earn-ings is complex because there are additional wo r k -related costs on the part of both on-site wo r ke r s(travel, costs of family care, etc.) and home-basedwo r kers (overhead for work place, utilities andequipment).

Street vendors

Street vending is a global phenomenon. In cities,towns, and villages throughout the world, millionsof people earn their living wholly or partly byselling a wide range of goods on the streets, side-walks, and other public spaces. With the advanceof modern retailing – fi xed retail operations,department stores, and malls – many expected thatstreet vending would go away. Yet today, in mostcountries of the world, street vending persists – andprobably has expanded – even where local regula-tions seek to ban or restrict it. It represents a feature

of traditional societies that has survived, adapted,or re-emerged in modern ones.

Who are Street Vendors?

Around the world, a large and, perhaps, grow i n gshare of the informal workforce operates on streets,sidewalks, and public parks, outside any enclosedpremise or covered workspace. This includes notonly those street vendors who sell goods but alsoa broader range of street wo r kers who sell servicesand produce or repair goods, such as: hairdressersor barbers ; shoe shiners and shoe repairers ; carwindow cleaners; tailors specializing in mending;b i cycle, motorcycle, van, and truck mechanics ;furniture makers; metal workers; garbage pickersand waste recyclers; headloaders and cart pullers;wandering minstrels, magicians, acrobats, andj u g g l e r s; beg gars and mendicants. In Ke nya, theS wahili term «Jua Kali» – which means «under theburning sun» – is the traditional name for theinformal economy. This is because so manyinformal activities, not just street trade, take placein the open-air under the burning sun.

E ven when used in the more narrow andprecise sense of informal traders who sell goodsfrom the street or in the open air, street vending isa large and diverse activ i t y : from high-incomevendors who sell luxury goods at flea markets tol ow-income vendors who sell fruits and veg e t a b l e salongside city streets. Those who sell a singleproduct or range of products as street vendors alsooften do so under quite different economic arrange-ments: some are truly self-employed and indepen-dent, others are semi-dependent (e.g., agents whosell products for firms against a commission),while still others are paid employees and fullydependent (see box presenting a typology ofstreet vendors).

Common Problems of Street Vendors

Street vendors are often viewed as a nuisance orobstruction to other commerce and the free flow oft r a ffic. Since they typically lack legal status andrecognition, they often experience frequent harass-ment and evictions from their selling place by localauthorities or competing shopkeepers. Their goodsmay be confiscated and arrests are not uncommon.The places where they work are often dirty andhazardous. Nevertheless, street vending may be theonly option for many poor people. Therefore, theright to vend – within reasonable limits orconstraints – should be considered a basiceconomic right (see box with summary of theBellagio International Declaration of Stre e tVendors).



Typology of Street VendorsStreet vendors are not a homogeneous group. They can be categorized or grouped according

to e.g. what types of goods they sell, where they trade, and from what type of premise as well asby their employment status, as indicated below. Also, for some street vending is full-time primarywork; for others it is a part-time secondary job.

Types of Goods: What do They Sell? Foodstuffs : fruit and vegetables Cigarettes and matchesCooked food Newspapers and magazinesSnacks and soft drinks Manufactured goods Candies and sweets Second-hand goodsIce cream and popsicles

Location of Work: Where Do They Trade? System of open-air markets in designated Railway stations, subway stations,

areas on designated days bus stops/lorry stationsConcentrations of vendors in particular areas: Construction sites

central business district or residential Sports complexesneighbourhoods Home

Street corners or sidewalksType of Premise: From What Do They Vend?

Baskets or bowls placed on the ground Bicycleor carried on the head or body Wheeled push-carts

Mats or cloths spread on the ground Wheeled stalls with display casesStool or table Porch-front or window displayPole over shoulder Fixed shed, stall, or kiosk

Employment Status: Are They Independent or Dependent? Independent self-employed: with and Semi-dependent workers:

without employees e.g. commission agentsDependent employees: paid workers for other

street vendors or for wholesale/retail traders

Bellagio International Declaration of Street VendorsIn 1995, re p resentatives from street vendor associations and activists, lawyers, and re s e a rc h e r s

working with street vendors from 11 cities around the world met in Bellagio, Italy to form an inter-national alliance (now called StreetNet) of street vendor associations and of organizations workingwith street vendors. The founders of the network drafted the Bellagio International Declaration ofStreet Vendors.

The Bellagio Declaration identifies the following common problems of street vendors aroundthe world:

• No Legal Status, No Right to Vend• Lack of Space or Poor Location• Restrictions on Licensing, Costs of Regulation • Harassment, Bribes, Confiscation, and Evictions• Lack of Services and Infrastructure• Lack of Representation or VoiceThe Declaration urges national governments to incorporate street vendors in economic policies

relating to trade, financial policies relating to micro-entrepreneurs, and social policies relating tothe working poor. The Declaration also urges city governments to incorporate street vendors inurban planning processes and urban policies and to promote institutional mechanisms for streetvendor associations to voice grievances, make demands, and resolve disputes with other urbanstakeholders.Source: Bellagio International Declaration of Street Vendors drafted by founding members from 11 countries of the inter-national alliance of street vendors, StreetNet, at a meeting in Bellagio, Italy, 1995.

Except in societies where gender norms restrictwo m e n ’s mobility, women account for a majorshare of street vendors. How eve r, with a fewnotable exceptions in mainly African countries,women traders are more likely than men traders tohave the more risky work situations, by: operatingfrom an open rather than a covered space; oper-ating from the street rather than a cart or a stall ;operating from an insecure or illegal space; tradingin perishable goods; generating a lower volume oft r a d e; working as commission agents or employ e e sof other ve n d o r s ; and not employing others towork for them. 1 0 C o n s e q u e n t l y, women ve n d o r salso tend to earn less than men vendors. 11

Size and Contribution of Street Trade

Despite their numbers and visibility, there aref ew good estimates of the number of streetvendors. Many population censuses and labourforce surveys do not contain a question on “placeof wo r k ” with relevant response alternatives. Or,if they do, the results are not routinely tabu l a t e do r disseminated. The international standardc l a s s i fication of occupations (ISCO-88) includesthree sub- groups: ‘stall and market salespersons’,‘street vendors and related wo r ke r s ’ ; and ‘shoecleaning and other street services elementary occu-pations’. How eve r, these categories are rarelypresented, either separately or combined, in offi c i a lstatistics.

Some of the special surveys that have provedeffective at enumerating street vendors have high-lighted an important feature of street trade thatm a kes it difficult to measure: namely, there is agreat variance in the number of street ve n d o r scounted depending on the time of day or the seasonof the year. The number of vendors can fluctuatefrom one season to the next, one day to the next,and even during a single day. This is because somevendors only sell in the morning, afternoon, orevening; some sell only on weekends; and otherssell only during certain seasons. Some may movefrom one location to another during the day,appearing settled at each; while others may changewhat they sell from one season, month, or day toa n o t h e r. Further, the same vendor family or unitmay have several different stalls at the same marke tor in different markets, confusing the question ofwhat is a vendor unit or bu s i n e s s . 1 2 One specialstudy attempted an estimation of the number of thebasic core of regular street vendors in Mexico Cityin the mid-1990s. The study estimated that therewere roughly 185,000 regular street vendors andan additional 90,000 occasional street vendors inM exico City (see box on street vendors inMexico City).

In principle, establishment censuses ands u r veys have the potential for capturing ormeasuring street vendors. The best approach wo u l dbe to use coordinated – or mixed – household and


Street Vendors in Mexico City, MexicoUntil recently, officials in Mexico City had data only on those forms of street vending that they

recognized, notably vending in public markets, various rotating markets, and recognized concen-trations of vendors in residential areas. In the 1990s, attempts were made to collect figures on acity-wide basis of registered ambulatory vendors with stalls of various kinds. But there have beenno efforts to collect official data on “non-tolerated” vendors, those without stalls and without regis-tration. In the mid-1990s, a study of the political economy of street vending in Mexico Cityattempted a conservative estimate of the diff e rent forms of street vending based upon re p o rt e dfigures and estimates. The study estimated that the core street vendors in the city operated from thefollowing types of locations and numbers of stalls :

C o n c e n t r a t i o n s: 1,500 stalls on the street or sidewalks (not counting 10,500 stalls in empty lots)Rotating Markets: 38,000 stallsAmbulatory: 67,248 stallsMetro stations: 5,000 stallsStreet corners/neighbourhoods: 10,000These estimates suggest that there were a total of 121,738 stalls in the city. In terms of the

number of street vendors, the study generated a conservative estimate of 185,600 individualspermanently inclued in full-time street vending in Mexico City. This figure represents only the basiccore of regular vendors in the city. It does not include vendors who occasionally sell in the streetor who add to the numbers during peak commercial seasons. The inclusion of such temporaryvendors would probably add, the study estimates, at least 50 per cent to the figure above.S o u rc e : John C. Cross. Informal Politics: Street Vendors and the State in Mexico City (Stanford, California, StanfordUniversity Press, 1998)

establishment surveys (as is recommended forcapturing the informal economy in general). It isn o t ewo r t hy that when national data collectiona c t ivities give priority to the measurement of streetvendors, as in the 2000 Labour Force Survey inSouth Africa, significant numbers of these wo r ke r sare identified. The 2000 Labour Force Survey esti-mated that there were 323,000 food vendors and122,000 non-food vendors in the country, whilet h e 1995 household survey estimated only2 , 0 3 8 vendors nationwide. While there has beensome “ r e a l ” increase in the number of streetvendors in South Africa due both to the relaxationof apartheid laws and to decreasing formal wo r kopportunities, much of the increase between 1995and 2000 is “statistical”: that is, due to improvedmethods for prompting respondents and trainingfieldworkers and coders. 13

In 10 developing countries, the estimated shareof street vendors in total non-agricultural employ-

ment ranges from less than 2 per cent (in CostaRica in 1997) to 9 per cent (in Guatemala in 2000)and is 4 per cent or more in six of the countries( Tunisia, Benin, Ke nya, Guatemala, Mexico, andVenezuela) (table 3.6).

In many cities or towns in the deve l o p i n gworld, especially in Africa, street vendors consti-tute a significant share of total employment in theinformal economy and street vending units consti-tute a significant share of total informal enterprises.For example, street vendors represented around3 0 per cent of those employed in the informaleconomy in one town of Niger in 19821 4 and 38 p e rcent of those employed the informal economy inthe five main towns of Guinea in 1987.1 5 In one cityin India, a recent survey found that street vendorsrepresent about 7 per cent of the estimated informalworkforce (see box on street vending inAhmedabad city, India). That study also foundthat, due to social norms that restrict wo m e n ’s


Table 3.6 Street Vendors in Ten Developing Countries: Estimated Number, Share ofNon-agricultural Employment, Proportion of Women

Number of street vendors Per cent of non-agricultural Per cent Womenlabour force

Tunisia (1997) 125,619 6 2Benin (1992) 45,591 5 81Kenya (1999) 416,294 8 33India (1999-2000) 3,881,700 3 14Turkey (2000) urban 255,000 2 3Brazil (1991) 1,445,806 3 30Costa Rica (1997) 13,085 1 18Guatemala (2000) 259,203 9 55Mexico (2000) 1,286,287 4 44Venezuela (1997) 318,598 4 32S o u rc e: Jacques Charmes, 2002 (personal compilation of the author on the basis of official labour force statistics and nationalaccounts). A subset of these data was published in ILO, On Measuring Place of Work ( Geneva, 2002).

Street Vendors in Ahmedabad, IndiaThe city of Ahmedabad in We s t e rn India was once known as the “Manchester of India”

because of its large textile industry. Over the past four decades, however, most of the 75 largetextile mills closed. Many of the 100,000 or more retrenched mill workers have turned to vendingand hawking. As a result, the number of street vendors has been going up steadily. However, untilrecently, there was no reliable estimate of the number of street vendors. In 1999, in collaborationwith the trade union SEWA (the Self-Employed Wo m e n ’s Association), the Gujarat Institute forDevelopment Research carried out a survey of the informal economy in Ahmedabad which wasspecifically designed to capture home–based and street–based work. The SEWA-GIDR surv e ysuggests that there were about 80,000 street vendors in the city, representing nearly 7 per cent ofthe estimated informal workforce. Street vendors were mainly men (90 per cent). Source: Jeemol Unni. Urban Informal Sector: Size and Income Generation Processes in Gujarat: Parts I and II (NationalCouncil of Applied Economic Research, Reports # 2 and 3 Delhi: April and May 2000).

m o b i l i t y, many women traders sell from theirhomes (see box in earlier section of this chapteron working at home in India).

In countries where statistics were available –six in Africa, three in Asia – informal traders,mainly street vendors, represent a very highproportion (73 to 99 per cent) of employment intrade and a significant share (50 to 90 per cent)o f gross domestic product (GDP) from trade( t a b l e 3 . 7 ). In South Africa, at current prices for1999, the informal economy contributed an esti-mated 26 per cent of the value added in trade: thehighest for any sector, followed by 18 per cent of

value added in both construction and communityservices. 16

In most of the countries where statistics wereavailable, women accounted for between 50 and9 0 per cent of informal traders and between 20 and65 per cent of the value added in informal trade.The notable exceptions were two countries – Indiaand Tunisia – where social norms restrict wo m e n ’smobility outside the home : the share of wo m e namong informal traders in these countries were12 and 8 per cent, respectively (table 3.7). Thesefigures reflect the predominance of women in tradein Africa and Southeast Asia and the restrictions on


Table 3.7 Size and Contribution of Informal Sector in Trade and Women Traders in Informal Trade

Informal Sector as a Share of : Female Informal Traders as a Share of :

Total Trade Total Trade Total Informal Total InformalEmployment value added Trade Employment Trade value added

AfricaBenin 99 70 92 64Burkina Faso 95 46 66 30Chad 99 67 62 41Kenya 85 62 50 27Mali 98 57 81 46Tunisia 88 56 8 4

AsiaIndia 96 90 12 11Indonesia 93 77 49 38Philippines 73 52 72 22Source: Prepared by Jacques Charmes, based on official labour force statistics and national accounts.

Street Vendors in Durban, South AfricaA 1997 census of street vendors in Durban, South Africa, counted 19,000 street traders: 57 p e r

cent of these were in the central business district (CBD) and 30 per cent were in high-density butnon-central areas (such as Umlazi township). Sixty-one per cent of the street traders were women;39 per cent were men. This survey counted both street traders who sold goods (78 per cent of thetotal) and street workers who provided services (21 per cent of the total). Of those who sold goods,seven out of ten sold food stuffs (fruits, vegetables, meat, and poultry); some sold new and usedclothing; and less than 1 per cent sold traditional medicines. Those who sold fresh produce weremainly self-employed, while those who sold clothes were most often employed by someone else.Nearly half sold only one type of product, one quarter sold two types of products, and one quartersold three or more types of products. Only a few (3per cent) sold goods or foods produced bythemselves.

The Durban study found some important gender diff e rences in street trading. Although fewvendors overall hired other people, slightly more men (13 per cent) than women (10 per cent) did.Among those hired, there were nearly twice as many women as men. Nearly all of the women,but only two-thirds of the men, were vendors only. More men (61 per cent) than women (54 percent) purchase from a large wholesale dealer, rather than from a small retail shop. About six outof ten of the transborder or immigrant street traders were men. Source: Francie Lund, Women Street Traders in Urban South Africa: A Synthesis of Selected Research Findings, Centre forSocial and Development Studies Research Report No. 15 (September 1998).

wo m e n ’s mobility in North Africa, Middle East,and South Asia.

Consider the case of Benin. A 1992 survey often major cities in that country found that streettrade constituted 81 per cent of all economic units,64 per cent of total employment, and 69 per centof urban informal sector employment. This surveyfound that women represented 75 per cent of streetvendors in these cities; and that women streetvendors accounted for 26 per cent of those in theurban informal sector and 24 per cent of the totalurban wo r k f o r c e .1 7 In Durban, South Africa wo m e nalso represent a greater share of street vendors than

men, and there are important differences betweenwo m e n ’s and men’s work in street vending ( s e ebox on street vending in Durban, South A f r i c a ).

Traditionally, the preparation and sale of foodhas been an important source of income forwomen. A study on street food vendors in the mid-1980s showed that in many countries wo m e nowned and operated street food enterprises and thatwomen were major contributors to many male-runbusinesses (see table 3.8). This is because menwho run street food enterprises generally dependon the labour of women working at home for manyof the products sold.18


Table 3.8 Street Food Enterprises by Women’s Involvement (per cent of all units)

City Owner or operator Enterprise with female assistants

Woman Man Couple Paid Unpaid

Bogor, Indonesia 16 60 24 5 33Chonburi, Thailand 78 22 – 13 31Ile-Ife, Nigeria 94 6 – 19 15Iloilo, Philippines 63 10 27 4 11Kingston, Jamaica 44 46 10 –- –Manikganj, Bangladesh 99 – 10 25Minia, Egypt 17 83 – 1 34Pune, India 13 87 – 33 51Ziguinchor, Senegal 77 23 –- – 25S o u rc e: I rene Ti n k e r. Street Foods: Urban Food and Employment in Developing Countries. (New York, Oxford University Press, 1997)

NOTES1 Planning Commission, Report of the Task Force on Employment Opportunities (Government of India, New Delhi. 2001).2 Alan Felstead and Nick Jewson, In Work, At Home. (New York, Routledge, 1999). 3 Martha Chen, Jennefer Sebstad and Lesley O’Connell, “Counting the Invisible Workforce”, World Development, Vol. 27, No. 3, pp. 603-610(1999) ; Alan Felstead, Nick Jewson, Annie Phizacklea and Sally Walters, A Statistical Portrait of Working at Home in the UK: Evidence fromthe Labour Force Survey, Economic and Social Research Council Research Programme on the Future of Work, Working Paper No. 2, (March2000).4 Jacques Charmes, Informal Sector, Poverty and Gender: A Rev i ew of Empirical Evidence. (Washington, D.C., The World Bank., 1998); S . V. Sethuraman, Gender, Informality and Pove r t y, (Washington, D.C., The World Bank, 1998) ; Felstead, et al., op. cit. ; Martha Chen andDonald Snodgrass, “M a n aging Resources, Activities, and Risk in Urban India: The Impact of SEWA Bank”, (Washington, D.C. USAID, 2001).5 ILO, “Decent Work and the Informal Economy”, Report VI, International Labour Conference, 90th Session, (Geneva, 2002). 6 Felstead, et al., op. cit.7 Jane Tate, “E very Pair Tells a Story”, Report on a survey of Homeworking and Subcontracting Chains in Six Countries of the European Union,(March 1996). 8 Jacques Charmes, “Is asking for ‘place of work’a pertinent and efficient way to better measure and understand the category of homeworkers,and more generally, outworkers in the labour force?” in ILO, On Measuring Place of Work, (Geneva: 2002).9 Linda N. Edwards and Elizabeth Field-Hendrey, “Home-based workers: data from the 1990 Census of Population”, Monthly Labor Review,(November 1996).10 Monique Cohen, with Mihir Bhatt and Pat Horn, “Women Street Vendors: The Road to Recognition, Seeds, No. 20, pp. 1-23, (New York,Population Council, 2000). 1 1 I b i d . and Martha Chen and Donald Snodgrass, “M a n aging Resources, Activities, and Risk in Urban India: The Impact of SEWA Bank”,(Washington, D.C., Management Systems International, 2001).12 T.G. McGee, Hawkers in Hong Kong, Center for Asian Studies Monographs and Occasional Papers, no. 17, (Hong Kong, University of HongKong, 1973); John C. Cross, Informal Politics: Street Vendors and the State in Mexico City , (Stanford, California, Stanford University Press,1998). 13 Debbie Budlender, Peter Buwembo, and Nozipho Shabala, “Country Case Study: South Africa”. The Informal Economy: Statistical Dataand Research Findings, (Geneva, International Labour Organization, 2001).14 Jacques Charmes, Le secteur non structuré à Niamey. Rapport d’une enquête. (Addis Abeba, Bureau international du Travail, Programmedes Emplois et des Compétences, 1982)

15 DGSI/PAGEN, Recensement des établissements dans les villes de Conakry, Kindia, Mamou, Labé, Kankan (Guinée). (Conakry, Ministèredu Plan et de la Coopération Internationale, Direction Générale de la Statistique et de l’Informatique, 1987).16 Debbie Budlender, et al. op.cit..1 7 C. Maldonado, Recensement des établissements en milieu urbain au Bénin. (Genève, PEESI/PNUD/BIT/INSAE, 1994); and Jacques Charmes,Situation et perspectives de la population active et de l’emploi au Bénin. (Cotonou, Ministère du Plan, de la Restructuration Economique et dela Promotion de l’Emploi, PNUD, 1996).18 Monique Cohen et. al., op.cit.


Why are statistics crucial?

What is clear from the statistics presented in thisbooklet is that the informal economy is far largerthan most people recognise. In all regions of thed eveloping world, informal employment comprisesnearly half or more – and in some countries morethan 75 per cent – of non-agricultural employ m e n t .In those developing countries that include agricul-ture in their estimates of informal employment, thesignificance of informal employment is generallyeven greater and can be as much as 90 per cent oftotal employ m e n t .1 In the United States andEurope, 25-30 per cent of the total workforce isi n part-time work, temporary work, and self-e m p l oyment, excluding other types of non-stan-dard work which are even more likely to beinformal in nature (such as casual day labour andvarious forms of contract work, including indus-trial outwork).

S e l f - e m p l oyment comprises 60 per cent ormore of informal employment (outside of agricul-ture) in all regions of the developing world: 60 percent in Latin America and Asia, 62 per cent inNorth Africa, and 70 per cent in sub-SaharanAfrica. In fact, self-employment represents nearlyone-third of total non-agricultural employ m e n tworldwide.

Informal wage employment is also significantaround the world. In the developing regions, wagee m p l oyment comprises 30 to 45 per cent ofinformal employment (outside of agriculture).Non-standard wage employment, much of whichis informal, is also significant in developed coun-tries. Part-time work alone, some of which isinformal, represents just under 15 per cent of totale m p l oyment in the OECD countries as a whole andmore than 20 per cent of total employment in eightof these countries.

What is also clear from the statistics is thatgenerally a greater share of the female workforcethan of the male workforce is in the informaleconomy. Moreover, although fewer women thanmen are economically active, women representt h e majority of wo r kers in several categories of

informal and atypical wo r k : notably, part-timework, temporary work, and homework. The find-ings presented here – although limited – alsosuggest that many people work at home or on thestreets. Ta ken together, home-based work andstreet trade represent 10-25 per cent of non-agri-cultural employment in developing countries andover 5 per cent of total employment in developedcountries.

In brief, these findings suggest that a larg eshare of the global workforce – and a signifi c a n tshare of the global female workforce – is employ e din informal enterprises and/or informal jobs.

Why should we care? As noted earlier, by defi-nition, those who work in informal employ m e n tare less likely than those who work in formale m p l oyment to enjoy wo r ker benefits and socialprotection. Although there are few official statis-tics on other characteristics of informal employ-ment, available data and research findings suggestthat informal wo r kers also earn less, on ave r a g e ,than formal workers; that the working poor, espe-cially women, are concentrated in the informale c o n o m y ; and that most low-income fa m i l i e ssecure their livelihood in the informal economy. 2

Available evidence also suggests that the informaleconomy creates large numbers of jobs, producesvast amounts of goods and services, andcontributes directly or indirectly to large shares ofexports. For many countries where the calculationcould be made, informal employment contributesabout a quarter of total GDP. In sum, the informaleconomy contributes to poverty alleviation and tothe total economy by producing a significant shareof total employment and GDP.

This booklet has presented the first statisticaloverview of the informal economy worldwide. Assuch, it is only a beginning. However, it suggestsimportant new directions – both substantive andoperational – for future work to further improvestatistics on the informal economy. In terms of howbest to proceed, it is important to build on thee fforts to improve statistics on the informaleconomy that contributed to this booklet. T h e s einclude:

4. IMPROVED STATISTICS FOR POLICY MAKING

■ A c t ive collaboration between the producersand users of statistics, including informedusers – both activists and researchers – whowere already deeply invo l ved in issues relatingto the informal economy;

■ Special efforts by countries such as India,M exico, and South Africa which have show nthe feasibility of collecting data on informale m p l oyment, including agriculture, and theusefulness of doing so ;

■ Technical expertise of the ILO has assistedmany countries in undertaking informal sectors u r veys and has played a lead role in deve l-oping the new broader conceptual framework;and

■ E fforts of the International Expert Group onInformal Sector Statistics, called the DelhiGroup, which has encouraged the deve l o p m e n tof the broader conceptual framework anddialogue between national statistical ex p e r t sand informed users.This booklet has highlighted limitations with

existing statistics. Few countries collect the statis-tics needed to estimate directly employment in theinformal economy as a whole and in its va r i o u scomponents. New data collection and compilatione fforts are required in countries, including :detailed questions in surveys and tabulations onkey indicators of self-employ m e n t ; informalaspects of non-standard wage employ m e n tincluding part-time wo r kers, temporary wo r ke r s( working on call or through temp agencies), casualday labourers, contract wo r kers, industrialo u t wo r kers (including homewo r ke r s ) ; and streetvendors and home-based workers.

Also, the country case studies have emphasisedthe importance of informal employment in agri-culture and the need to address it in internationalwork on the informal economy.

I m p r oved statistics would help to focus theattention of policy makers on the contributions ofthe informal economy to economic growth and onthe linkages between informal employ m e n tarrangements and pove r t y. More specifi c a l l y,i m p r oved statistics on the informal economy wo u l dserve :■ to increase the visibility of those who work in

the informal economy, especially the leastvisible and most neglected workers ;

■ to advance understanding of the informale c o n o m y, including its contribution toeconomic growth and its links with poverty;

■ to highlight the gender dimensions of theinformal economy; and

■ to inform the design of appropriate policies andprogrammes for those who work in theinformal economy and to help monitor andevaluate the impact of these (and other) poli-cies and programmes on those who work in theinformal economy.

Objectives for Future Work

The overall objective of future work on statisticson the informal economy should be to ensure thatestimates on the size, composition, and economicvalue of the informal economy, disaggregated bysex and also, ideally, by age, are incorporated intoofficial statistics at national, regional, and interna-tional levels in a systematic and on-going basis. A nimportant aspect of this is to promote the compa-rability of data across countries and over time. T h es p e c i fic goals should be to improve methods forthe collection and compilation of statistics ; tostrengthen the capacity of countries to collect thesedata; to develop and manage a pooled data base;to promote on-going compilation and analysis ofexisting statistics ; to generate national, reg i o n a l ,and global estimates ; and to disseminate improve dconcepts, methods, findings, and estimates.

Plan of Action

To pursue these objectives, the joint action ofnational, regional, and international statisticalservices, of relevant governmental and UnitedNations agencies, and of activists and researchersworking on the informal economy will be required.The respective areas of responsibility for thesedifferent stakeholders will need to vary dependingon the actions being undertaken. The division oflabour between different stake-holders will need tot a ke into account existing inter- a g e n cy agreementswhereby the ILO Bureau of Statistics is respon-sible for labour statistics and the UN StatisticsD ivision is responsible for national accounts statis-tics and overall coordination of the work. T h eactions to be taken should cover the following:

Concepts and Methods to Improve DataCollection and Tabulation

■ I m p r ove statistical concepts and methods t ogenerate more accurate and comprehensivestatistics on the informal economy: that is,informal employment both in and outsideinformal enterprises, and including in agricul-ture.


■ Encourage the active collaboration of statisti-c i a n s in national, regional and internationalagencies with data users in identifying dataneeds and in the design and testing of appro-priate methods.

■ Give special attention to improving measuresof those engaged in informal employmentoutside informal enterprises, including manytemporary wo r kers and part-time wo r kers aswell as most industrial outwo r kers and mostcasual wo r kers with no fi xed employ e r. T h i sshould involve further methodological effortsrelated to cross-classifying these forms of non-standard or atypical work with key criteria ofinformal employ m e n t : namely, security ofcontract, presence/lack of wo r ker benefits, andsocial protection coverage.

■ Give special attention to improving measuresof those working at home, on the street, or inthe open air. This should invo l ve furthermethodological efforts related to the “place ofwork” variable and further efforts to collectand tabulate these data in national data collec-tion activities. While the “place of work” vari-able is not directly related to improving statis-tics on the overall size of informale m p l oyment, it is an important variable inanalysing the composition and characteristicsof informal employment.

■ I m p r ove the measurement and tabulation ofvariables relating to other key aspects of workin the informal economy : for example, season-ality and other temporal patterns ; and multipleand secondary occupations.

■ Ensure that the special methodological issuesrelated to the collection of data on diff e r e n tsub-groups of the informal economy – womenand men, age groups, rural-urban residence,and socio-economic strata – are considered inthe design of all data collection and tabulationactivities.

■ P r ovide t e chnical assistance to countries tocollect statistics on the informal economy ift h ey do not already do so or to improve thequality of statistics they already collect. Two

o b j e c t ives of technical assistance would be i) t oimprove the international comparability of thedata as a basis for preparing regional andglobal estimates of the informal economy; andii) to improve the measurement of wo m e n ’sactivities in the informal economy.

Preparation and Dissemination of Policy-Relevant Statistics on the Informal Economy

■ Develop, maintain and update at regular inter-vals a c o m p rehensive database on the informale c o n o m y, incorporating existing data oninformal employment and total employ m e n tfrom as many countries as possible. A ny futurework should link data on informal employ m e n twith data on total employment to allow for acomplete statistical picture of the total wo r k-force, including the share of informal employ-ment (and its constituent parts) in total employ-ment.

■ Specify key indicators of informal employmentfor which data should be routinely collectedand tabulated for the database and for gener-ating estimates. Key indicators relating to theinformal economy would include the size,composition, and contribution to GDP.

■ Generate national, regional and global esti-m a t e s at regular intervals on key indicatorsrelated to the development of policies andprogrammes on informal employment anddisseminate the estimates in user- o r i e n t e dformats. The preparation of regional and globalestimates needs to be based on a comprehen-sive programme in which new work is under-taken rather than an ad hoc compilation of theavailable data.

■ Analyse and publish data in u s e r- o r i e n t e dre p o r t s. Data to be published should includethe size, composition, and contribution of theinformal economy and special topics such asthe links between informality, poverty andgender and the size, characteristics and contri-bution of specific groups of wo r kers. T h eformat of these reports should be accessible toa wide group of users.

Improved Statistics for Policy Making 59

NOTES1 M a ny countries define and measure informal employment outside agriculture. The size and significance of informal employment are far higherin countries that define and measure informal employment both within and outside agriculture (see Main Findings and country case studies inChapter 3).2 Jacques Charmes, Informal Sector, Poverty and Gender : A Rev i ew of Empirical Evidence. (paper prepared for the World Bank, 1998) ; S.V. Sethuraman, Gender, Informality and Poverty, (paper prepared for the World Bank, 1998). Both papers are available at www.wiego.org.

Africa

Northern AfricaAlgeriaEgyptLibyan Arab JamahiriyaMoroccoTunisiaWestern Sahara

Sub-Saharan AfricaAngolaBeninBotswanaBurkina FasoBurundiCameroonCape VerdeCentral African RepublicChadComorosCongoCôte d’IvoireDemocratic Republic of theCongoDjiboutiEquatorial GuineaEritreaEthiopiaGabonGambiaGhanaGuineaGuinea-BissauKenyaLesothoLiberiaMadagascarMalawiMaliMauritaniaMauritiusMozambiqueNamibiaNigerNigeriaReunionRwandaSao Tome and PrincipeSenegalSeychelles

Sierra LeoneSomaliaSouth AfricaSudanSwazilandTogoUgandaUnited Republic of TanzaniaZambiaZimbabwe

Latin America and theCaribbean

CaribbeanAntigua and BarbudaArubaBahamasBarbadosCubaDominicaDominican RepublicGrenadaGuadeloupeHaitiJamaicaMartiniqueNetherlands AntillesPuerto RicoSaint Kitts and NevisSaint LuciaSaint Vincent and theGrenadinesTrinidad and TobagoUnited States Virgin Islands

Central AmericaBelizeCosta RicaEl SalvadorGuatemalaHondurasMexicoNicaraguaPanama

South AmericaArgentinaBoliviaBrazilChile

ColombiaEcuadorFrench GuianaGuyanaParaguayPeruSurinameUruguayVenezuela

Asia

Eastern AsiaChina

Hong Kong Special Administrative RegionMacao Special Administrative Region

Democratic People’s Republic of Korea

MongoliaRepublic of Korea

South-eastern AsiaBrunei DarussalamCambodiaEast TimorIndonesiaLao People’s DemocraticRepublicMalaysiaMyanmarPhilippinesSingaporeThailandViet Nam

Southern AsiaAfghanistanBangladeshBhutanIndiaIran (Islamic Republic of)MaldivesNepalPakistanSri Lanka

Central AsiaKazakhstan*

Annex 1 Countries, areas and geographical groupings

ANNEX

Annexes 61

Kyrgyzstan*Tajikistan*Turkmenistan*Uzbekistan*

Western AsiaArmenia*Azerbaijan*BahrainCyprusGeorgia*IraqIsraelJordanKuwaitLebanonOccupied Palestinian TerritoryOmanQatarSaudi ArabiaSyrian Arab RepublicTurkeyUnited Arab EmiratesYemen

Oceania

American SamoaFijiFrench PolynesiaGuamKiribatiMarshall IslandsMicronesia (Federated States of)Nauru

New CaledoniaPalauPapua New GuineaSamoaSolomon IslandsTongaVanuatu

Developed Regions

Eastern EuropeAlbania*Belarus*Bosnia and Herzegovina*Bulgaria*Croatia*Czech Republic*Estonia*Hungary*Latvia*Lithuania*Poland*Republic of Moldova*Romania*Russian Federation*Slovakia*Slovenia*The former Yugoslav Republic of

Macedonia*Ukraine*Yugoslavia*

Western EuropeAndorraAustria

BelgiumDenmarkFinlandFranceGermanyGreeceIcelandIrelandItalyLiechtensteinLuxembourgMaltaMonacoNetherlandsNorwayPortugalSan MarinoSpainSwedenSwitzerlandUnited Kingdom of Great Britain

and Northern Ireland

Other Developed Regions AustraliaBermudaCanadaJapanNew ZealandUnited States of America

*included in the category of ‘countries intransition’


Annex 2 Self-employment as percentage of non-agricultural employment

Region1980/1990 1990/2000

All persons Women Men All persons Women Men

WORLD 26 28 25 32 34 27

Developed regions 13 11 14 12 10 14

Eastern Europe 18 15 21 9 7 11Bulgaria 0 0 0 – – –Croatia 8 6 9Czech Rep. 13 9 16Hungary 6 6 5 12 12 13Lithuania 63 53 73 – – –Poland 4 2 6 12 9 13Romania 6 4 8Russia 5 – –Slovakia 5 5 4Slovenia 9 6 12

Western Europe 12 10 14 14 11 17Austria 7 8 7 8 7 9Belgium 14 14 13 – – –Denmark 7 5 9 – – –Finland 6 5 8 9 6 12France 10 9 11 11 10 12Germany 8 8 9 10 7 12Greece 31 20 35 31 24 35Italy 20 17 21 26 22 28Ireland 18 10 22 – – –Netherlands 9 9 9 9 7 10Norway 8 6 10 9 7 10Portugal 15 10 18 19 15 21Spain 21 20 22 19 16 20Sweden 3 1 5 9 5 13United Kingdom 8 4 10 13 8 17

Other developed countries 10 9 11 11 9 12Australia 14 11 16 9 7 11Canada 5 3 6 8 8 9Japan 18 22 15 14 16 12New Zealand 8 5 10 16 11 21USA 6 4 7 7 6 8

Africa 44 58 37 48 53 37Northern Africa 23 23 22 31 38 28Algeria 13 5 13 28 35 27Egypt 24 4 27 25 12 27Morocco 36 44 33 40 46 37Tunisia 21 40 15 30 61 20Sub-Saharan Africa 50 69 41 53 57 40Benin 89 96 80Botswana 7 8 6 13 15 11Burkina Faso 78 93 65 70 – –Burundi 36 42 34

Annexes 63

Region1980/1990 1990/2000


Central Africa Republic 58 86 44 – – –Chad 72 95 57Comoros 36 60 32 61 77 –Congo 39 61 30 – – –Ethiopia 68 79 52Gambia 95 97 95 52 51 52Ghana 67 86 40 – – –Guinea 72 83 42Guinea Bissau 48 66 –Kenya 61 79 –Liberia 41 65 33 – – –Mali 78 90 66 84 – –Mauritius 17 12 19Mozambique 20 31 19 47 61 42South Africa 17 19 7Sudan 41 22 46Togo 68 89 50 – – –Zaire 52 82 42 – – –Zambia 18 43 12 – – –

Latin America 29 30 27 44 54 35

Central America 30 32 27 40 54 29Costa Rica 18 10 22 33 42 28El Salvador 47 68 28Guatemala 45 65 29 – – –Honduras 52 69 35Mexico 42 44 41 39 61 26Panama 15 9 18 30 31 30

South America 29 29 29 43 51 38Argentina 42 61 32 29 26 31Bolivia 68 95 46Brazil 21 16 23 35 40 32Chile 20 15 22 29 35 25Colombia 31 32 31 49 58 42Ecuador 40 34 42 49 44 52Uruguay 22 19 24 – – –Venezuela 27 28 26 44 56 37

Caribbean 27 28 25 55 67 43Cuba 1 1 1 – – –Dominican 24 18 28 45 54 40Haiti 57 64 45 65 80 46

Asia 26 24 26 32 32 30

Eastern Asia 23 20 24 18 26 16China 15 – –Hong Kong 13 9 15 10 20 3Republic of Korea 32 32 32 30 32 29


Region1980/1990 1990/2000


South-eastern Asia 34 38 31 33 40 24Cambodia 49 66 39Indonesia 56 70 50 51 60 –Malaysia 23 21 24 17 17 17Philippines 34 41 28Singapore 16 10 20 13 14 13Thailand 40 51 32 36 40 –

Southern Asia 40 38 40 50 55 48Bangladesh 53 53 53 75 83 73India 44 55 42 37 41 38Iran 35 32 36 37 39 37Nepal 59 80 49Pakistan 51 40 51 43 34 44Sri Lanka 17 9 19

Western Asia 13 6 14 24 10 22Bahrain 10 2 11 5 1 6Iraq 14 11 15 – – –Jordan 24 6 –Kuwait 6 0 8 – – –Lebanon 34 16 39Oman 4 2 4Qatar 2 0 2 – – –Syria 25 15 26 29 15 31Turkey 24 14 25 28 15 30United Arab Emirates 8 2 9 – – –Yemen 41 17 –

Source: Data prepared by Jaqcues Charmes .

Note: Data were compiled from population censuses, labour force surveys and from a few living standard surveys. Most populationcensus data were in the Yearbook of Labour Statistics: Retrospective Edition on Population Censuses, 1945-89 (International LabourOrganization, Geneva 1990). The ILO Yearbook stopped publishing the required table (2A) which cross-classifies the economicallyactive population by industry, status and sex. ILO is preparing a special issue of the Yearbook including this specific table and pro v i d e daccess to the new unpublished data now available from countries. The United Nations Economic and Social Commission for WesternAsia (ESCWA) Statistics Division, Labour Force Statistics Database CD-ROM version 3.0 (Amman Jordan, 1997) was also used fordata for the Arab countries.

women and men in the informal economy : a statistical picture · picture, using the available data,...

Documents