1.2.3 Surveys and data collection in a historical perspective
1.2. A brief history of 40 years of conceptualisation and data collection on the informal economy
1.2.3. Surveys and data collection in a historical perspective
The main sources of data are the most recent national labour force surveys and/or the mixed (household/establishment) surveys. However, many of the published reports are not always available and where they are available, they may not contain the required classifications and tabulations.In some countries, the reference to the concepts of informal employment or informal sector is not even mentioned. In these cases in particular, the main source of data are the ILO questionnaires sent in 2011 by the ILO Bureau to all statistical offices of member countries (developing countries and transition countries). It requested that the national offices compile detailed tables with statistics on employment in the informal sector and informal employment, with a special table on metadata conveying the information of survey coverage and definitions of concepts.
The detailed sources and specificities of definitions according to national circumstances can be found in Charmes (2011).
The enterprise-based approach taken during much of the1970s and 80s makes this period stand out more than others. The strong emphasis on this approach is perhaps not so surprising when considering that during these decades it was deemed necessary to construct all national accounts and conceptualise GDP with data collected on earnings and production. Economic and door-to-door censuses of establishments were regularly conducted and followed by sample surveys. It is also during thisperiod that theadapted and sophisticated designs of questionnaires were tested for the measurement of production,as seenfor instance, in Tunisia between1976-1982,wheredirect reportingwas often underestimated by half compared with other controlled methods.
The census approach of activities, however, even when extended to mobile (non-sedentary) vendors failed to capture the bulk of home-based workers or rather outworkers - those workers who do not perform their activities in the premises of an enterprise and are not enterprise-based.
It is for this reason that there was a change of methodological paradigm that occurred from the end of the 1980s through the 1993 International Conference of Labour Statisticians,which defined the concept of informal sector. It was during this period when the first mixed household-establishment surveys were conducted in Mali (1989) and in Mexico (1991), just before the 1993 ICLS recommendation proposed this type of survey as the most appropriate for capturing all the diversity of informal sector activities. Many countries conducted such surveys at national level (India, 1999-00; Tanzania, 1991; South Africa, 2002; Cameroon, 2005; Morocco, 2007; among others) or at capital city or urban levels (the series of 1-2-3 surveys in the eight francophone countries of West Africa as well as in Cameroon and Madagascar) during the 1990s and early 2000s.
Asia followed in the second half of the 2000s with mixed surveys in Bangladesh, the Philippines and Indonesia as well as Cambodia, Mongolia and Armenia. The decades of the 1990s and 2000s are thus those of mixed surveys.
During the same period, efforts started to include adapted questions or even short sections in the questionnaires of regular household surveys (labour force surveys or living conditions surveys) in Latin America and in Asia (Pakistan, Thailand), while the LSMS questionnaires (and the surveys of the same type, for instance the GLSS in Ghana), as well as the “integrated” or “priority” surveys on living conditions of households, introduced a section for capturing the activities of own-account and employers’ enterprises.
With the 2002 International Labour Conference (ILO, 2002a and b) and the 2003 ICLS (ILO, 2003), the pendulum swingsback to emphasise the individual-based definitions. Efforts are made to gather information on the type of contracts and social protection for the paid employees and the benefit of some kind of social protection for all the workers and the whole population more generally through household surveys.
To sum up, one can say that the first two decades (1970s and 1980s) were decades of establishment censuses and surveys, which remains a continued concern for national accounts purposes. This period allowed for the obtaining ofbetter knowledge of the upper tier of the informal sector (i.e.micro and small enterprises or MSEs).
The following decade (1990s) until the beginning of the 2000s wasthe decade of mixed surveys, achieving the requirement of accumulating knowledge on the characteristics of the various components of the informal sector including the lower tiers and employment creation.
Finally the last decade (2000s) saw the rise of the household surveys as the main vehicle of data collection on informality. This occurred firstly because these surveyshad conveniently been the first stage of the mixed surveys, secondly because they have often become regular – if not permanent (annual or even quarterly) – and thirdly because they can accommodate a special section or module to informality in its broad sense (informal employment and informal sector).
It is not yet obvious whether the 2010s will see arepetition of mixed surveys at national level, although there are some signs that it is heading in this direction (Madagascar, Niger, Cameroon, RD Congo).