[ Table of contents ]   [ List of papers ] Health Impact of Occupational Risks in the Informal Sector in Zimbabwe Chapter 3 Survey of occupational risks and health impacts in the informal sector in Zimbabwe

3.1 Objectives and methods of the survey

The survey of occupational risks and health impacts in the informal sector in Zimbabwe was carried out to

1. examine the pattern of occupational risks in the urban informal manufacturing sector and the rural peasant (informal) sector
2. identify the rate and pattern of injury in the urban and rural informal sectors (as above), the major agents of accidents and the extent of compensation coverage of such injury, using the same definitions of compensable injury as applies in the formal sector
3. examine the levels and patterns of major occupational illnesses in the informal sectors above
4. obtain proxy information on work related fatalities in the informal sectors above
5. compare the information on rates and patterns in the informal sector with information for the equivalent sectors in the formal sector and with available findings from household / worksite surveys carried out elsewhere.

The study population was drawn from

• the urban informal sector, viz: urban home based enterprise and informal production sites,
• the rural informal sector, viz: peasant agriculture and rural informal sector production sites.

The survey covered rural peasant farming and urban production, including other production activities taking place in the informal sector rural farming and urban areas. The survey did not aim to include mining, transport and contruction, although these are also important infornmal sectors, as the financial resources did not cover their inclusion. They will need to be covered in a later stage.

The study areas were Harare (urban) and Murewa district (rural). The production sites were sampled on the basis of their size so that a census of activities in a defined site could be carried out. These were Magaba in Harare and a Murewa rural production site. The household samples covered two Central Statistical Office (CSO) enumeration areas from the national master sample in each of the rural and urban areas (Murewa rural and one each in Highfields and Mbare urban). The households included were identical to those included in the CSO household survey programme in those enumeration areas and were a census of the enumeration areas.

An estimated injury rate of 2% was used to estimate the sample size, based on official reported data in the formal sector (NSSA 1995). With 95% confidence intervals the sample size required for an error of 1% was calculated to be 753. Hence a sample of 760 was chosen to enable detection of a 2% injury rate with a +/- 1% standard deviation (using a 95% confidence interval). It was thus intended that 760 workers would be covered in each of the rural and urban areas, or 1 520 workers in total. In fact, 892 workers were included in the urban sample and 678 in the rural sample, with a total of 1 585 respondents.

 

Informal sector employees were defined as those working in registered and licensed enterprise, employing less than 10 people, not covered by formal collective bargaining mechanisms and excluding full time students and unemployed people. Field workers were told that if there was doubt the worker should be included and the doubtful cases would be identified in the analysis.

All people working in the household or worksite, except those working in paid employment in the formal sector were included. Where respondents were not present field workers returned to interview them at another time. Where they were not capable of responding (ill, absent, too young (<12 years) or too old), the non response was recorded with the area and cause, for analysis of loss to the sample.

The questionnaire for the survey was developed using existing surveys of occupational risks and injury, some linked to labour market and other household surveys. The questionnaire was particularly drawn from survey questions used in UK, South Africa and Pakistan. In addition, the survey harmonised its labour market section and definitions with the Zimbabwe CSO labour market survey questionnaire and the work environment hazards section was based on a checklist of work hazards used in a Zimbabwean survey of 100 formal sector workplaces, carried out by the author and the Zimbabwe Congress of Trade Unions in 1993. The questionnaire was widely circulated for comment and adjusted after feedback from ILO (SEC HYG and STAT), the Zimbabwe CSO and input from a pilot field test carried out in urban areas. The final questionnaire used is shown in Appendix A. One objective of the survey was to field test an instrument that could be used as a module attached to the routine household survey programme. This was discussed with the relevant section of the CSO, and further comment on the form used in this regard will be discussed in Section 5.2.

The field workers were given training on the research instrument and the occupational health concepts it is based on by two trade union occupational health officers (S Chaikosa and G Marembo) who were part of the team, by an experienced safety engineer (W Husberg) and the author, an epidemiologist. After field testing of the questionnaire further training was carried out, including training by enumerator supervisors of the CSO on household survey techniques. The latter also formed part of then research team. Written guidelines on how to respond to questions were prepared for field workers. Quality control of questionnaires was carried out by the field work co-ordinators and by cross check of a subsample of questionniares by the author.

Data entry, cleaning and analysis was done on SPSS at the University of Zimbabwe Statistics department (E Marinda and H Chipoyera). It is intended that before the final draft of the report is prepared a meeting will be held with the field workers and CSO Zimbabwe to discuss the findings and in particular to discuss areas that require further interpretation, explanation of variance or unusual findings etc.

Ethical guidelines were followed. The stakeholders, and in particular the state authorities responsible for occupational health were consulted on the design and questionnaire for the study. The full report will be tabled at a meeting of the national tripartite council on occupational health to identify potential areas of follow up intervention. Community leaders were briefed about the survey in the areas sampled prior to its commencement and feedback will be given to them on the results and their implications. Individuals were offered the voluntary option to respond. Confidentiality of individual information is incorporated and was noted. The survey does not involve any direct medical surveillance procedures.

3.2 Survey results

A total of 1 585 workers were interviewed in the survey, subdivided as shown in Table 3.2.1 below. Excluding the missing data, 57% of the sample was urban, 43% rural, 71% drawn from household data and 29% from production sites.

(*) Total where area stated = 1570. In a further 15 interviews area was not stated.

The demographic characteristics shown in Tables 3.2.2a and Table 3.2.2b indicate that in urban areas, the greater share of workers are relatively young (30% between 20-24 years old, 21% 24-29 years old), while in rural areas the median age is even lower (16% 15-19 years and 15% 20-24 years old). This differs from prior 1985 and 1991 surveys which found a mean age of 37-38 years (Mhone et al 1985; USAID 1991). Notably these surveys did not cover rural areas or household production where age levels are lower.

The agricultural sector has a younger median age than other sectors, while in manufacturing the range is shifted slightly towards older age groups. The young age group of workers signifies the large number of school leavers entering the informal sector, and the survey reflects the 1992 Census findings that informal sector employees are younger than formal sector (CSO 1994). In relation to occupational health the young age range signals job inexperience which may increase accident risk and interactions between occupational and reproductive health.

Indicator	URBAN H/HOLD	URBAN PRODN	URBAN TOTAL	RURAL H/HOLD	RURAL PRODN	RURAL TOTAL	ALL TOTAL
Age in yrs < 15 15-19 20-24 25-29 30-34 35-39 40-44 45-49 50-59 60+	6 1 66 13 143 28 113 22 70 14 46 9 37 7 14 3 243 3 12 2	1 0.3 42 11 126 34 76 20 50 13 31 8 17 5 11 3 10 3 8 2	7 1 108 12 269 30 189 21 120 13 77 9 54 6 25 3 23 3 20 2	58 10 88 15 86 14 69 12 52 9 31 5 41 7 25 4 77 13 74 12	0 0 17 22 17 22 7 9 9 12 12 16 8 10 4 5 2 3 1 1	58 9 105 16 103 15 76 11 61 9 43 6 49 7 29 4 79 12 75 11	69 4 216 14 373 24 267 17 184 12 121 8 104 7 54 3 102 6 95 6
SEX: Male Female	233 45 284 55	302 82 68 18	535 60 352 40	271 46 324 54	44 59 31 41	315 47 355 53	856 54 715 46
Schooling < Grade 1 Grade 1-7 Form 1-6 > Form 6	0 0 171 34 320 63 15 3	0 0 89 24 258 70 19 5	0 0 260 30 578 66 34 4	0 0 341 62 200 36 12 2	0 0 37 49 36 47 3 4	0 0 378 60 236 38 15 2	56 4 644 41 822 52 49 3

In the urban and rural household surveys the majority of workers were female, while in the production sites the majority were male. Urban workers were more predominantly male than rural, and the total male / female ratio of 1,2:1 is closer to parity than the ratio in the formal sector, where the ratio is closer to 3,2:1 (CSO 1994). This is consistent with prior informal sector surveys which found that 57% of informal sector workers were female (Ndoro 1996). The highest proportion of male workers is found in the urban production sites and in manufacturing (79% male). The preponderance of female labour in household surveys signals the likelihood of women being involved in poorer quality jobs in home based enterprises, often not visibly recognised as work and linked with their existing domestic workload. Consistent with this women are also primarily involved in agriculture, retail and services.

The workers in the survey were relatively well educated. Surprisingly, none of the surveyed workers had less than grade 1 education. This partly reflects the fact that universal primary education has been in place since independence, but is surprising that even in the older age groups, all had at least entered primary education. If literacy is classified as entry into grade 3, then 97% of the workers were functionally literate. Over half of the workers had completed primary education, with 30% more reaching post primary levels in urban than in rural areas. Urban workers had significantly higher levels of education than rural workers (Chi Sq, p<0,001). The relatively high level of literacy and education indicates that there is potential for information flow through the informal sector using written and other media, and that educational status should not be a major limiting factor for the uptake of occupational health information.

Tables 3.2.3a and b show information on the employment status of the workers interviewed. In urban areas and the rural production site, the greater share of workers are self employed (57% and 65% respectively), while in the rural households the greater share are self employed peasant farmers (42%). Urban areas and rural production sites also have a relatively large share of casual, contract or seasonal employees employed by small enterprises (14% and 27% respectively). Excluding the agricultural sector where 42% of workers are peasant farmers, self employed workers comprise the larger share of other sectors of employment (43% to 65%). The highest levels of more formal forms of employment exist in small manufacturing enterprises (18% permanent employees and 13% casual or contract employees). Unpaid family work is relatively frequent in rural and urban households (21% and 9% workers respectively). Such contracts are amongst the most difficult to monitor and regulate.

Workers are employed for an average of 11,1 months a year, shortest in rural production sites (10,2 months) and longest in rural households (11,33 months). The retail sector had the shortest periods of activity in the year, and the agricultural sector the longest, although the variation was not significant. Workers reported working an average of 6,1 days a week, and 8,7 hours a day, or a working week of 53,1 hours (See Tables 3.2.3a and b). Comparing the informal sector working hours with the maximum statutory hours of work in the formal agricultural and manufacturing sectors, the average monthly hours of work in informal sector agriculture is less than formal sector statutory limits, but the hours in informal sector manufacturing exceeds the statutory limits by 17 hours ^{(Endnote 5)}. This extends the exposure of urban informal sector workers to work hazards and to fatigue due to overtime work.

INDICATOR	Agriculture	Manufacturing	Retail	Service/Other
JOB STATUS: permanent employee casual/contract/seasonal employer peasant farmer own account worker-other unpaid family work unemployed student homemaker apprentice /trainee retired other	18 3 11 2 2 0.3 253 40 99 17 129 22 1 0.2 59 10 18 3 2 0.3 3 1 1 0.2	60 18 44 13 23 7 1 0.3 19 56 17 5 1 0.3 1 0.3 1 0.3 0 0 0 0 1 0.3	33 8 63 16 5 1 0 0 265 65 26 6 1 0.3 3 1 0 0 2 0.5 1 0.3 6 1	23 14 32 20 5 3 1 1 69 43 10 6 1 1 3 2 10 6 0 0 0 0 6 4
<6 months worked/yr >7 months worked/yr	20 3 577 97	19 6 320 94	35 9 370 91	7 4 153 96
<3 days worked/yr >4 days worked/yr	22 4 569 96	12 0.3 335 99	3 1 399 99	4 2 155 98
<8 hours worked/day >9 hours worked/day	330 56 261 44	72 21 264 79	94 24 308 77	102 64 57 36

The distribution of jobs by area and sector is shown in Tables 3.2.4a and b overleaf. This table reflects the main economic activity carried out in the 12 months prior to the survey. The top five areas of activity and the proportion of workers engaged in them are shown in Table 3.2.5 overleaf.

The distribution of sectors is not surprising: manufacturing and retail predominates in urban areas, while agriculture and retail trade predominate in rural areas. Urban activity mainly centres around production of household goods for domestic markets (furniture, building goods), sale of food products and household items and services such as transport and hairdressing. ('Other' thus refers mainly to the service sector).

Rural activity is centred around farming (and sale of farm products) with servicing of domestic food, building and hardware needs. It is evident that the sector produces goods and services for the domestic, local market. It has been raised elsewhere that the purchasing power for these goods stems mainly from formal sector wages (Mhone 1996). These areas of employment carry risks which, from formal sector reported data, are shown in the risk exposure matrix in Table 3.2.6 overleaf. How do these risks compare with those found in the informal sector? This is discussed later in this section.

Skills levels of the workers interviewed are also shown in Tables 3.2.4a and b. While the majority of workers are semiskilled or unskilled (92% of workers), skills levels are higher in urban than in rural production, and higher in manufacturing and services than in agriculture or retail sectors. There is little training in the informal sector and skills are often acquired through in service forms of training. The possibility exists for including occupational health information in this exchange of knowledge, but it means that those who transmit skills, such as through apprenticeships, should also know and communicate information on assessment and reduction of occupational risk. It would need to be determined whether such knowledge exists and if not how it can be developed or enhanced.

Most workplaces are neither registered nor licensed (79%), particularly in the sample from urban households (95%) and rural households (84%). Production sites do have a higher proportion of workplaces that have licenses (urban 34% and rural 43%), these being issued by the local authority as a condition of their location at these sites.

The small proportion of such licenses reported in household surveys indicates that many home based enterprises are operating without licenses, or as an informal outreach of domestic work. The licensing system provides an opportunity for communicating information on occupational health and compensation provisions to informal sector operators in production sites, but reaching unlicensed workplaces, particularly those that operate from home, is more difficult. While the majority of urban home based activities relate to sale of goods, in both urban and rural areas production of furniture and service activities may involve hazards that require information on risk reduction.

(*) unless grouping is too small to include

(c) Based on injury reports to the National Social Security Authority (NSSA 1995)

Tables 3.2.7a and b also show the workforce and worksite sizes^{(Endnote 6)}. The total workers in the workplaces covered in the survey is estimated at 4 804^{(Endnote 7)}. The average size of the workforces in rural and urban informal sectors is predictably small, 2,57 in urban areas, 3,75 in rural areas, and 3,03 overall. This is consistent with the findings of earlier surveys (ILO/SATEP 1985; USAID 1991). Urban production sites and rural households have the larger number of workers per workplace, with 3,02 workers per workplace in the former and 4,01 in the latter.

Reaching workers in the informal sector is thus further complicated by the small number of workers per workplace or employer. The relationship used in the formal sector of relying on the employer is less easily applied in the informal sector, particularly given the fact that the 'employer' may not even be present much of the time. More community based approaches may be more relevant in this sector, reaching employees through area based / local community structures. Services that support the informal sector as a whole, such as in relation to credit support, marketing, health and hygiene provide such useful conduits for occupational health information and inspection.

Worksite sizes are either very small (19% less than 10m²) or very large (48% more than 70m²). The latter mainly covers activities involving public places, such as vending, transporting and farming, Predictably large worksites are found in rural households, where farming is carried out, and urban households (where greater shares of vending and services are carried out), while smaller worksites were found in rural and urban production sites. Small and concentrated worksites in production areas make such workplaces more accessible for surveillance and support. However the mobility and poorly delineated worksites of an equally large share of workers in the farming, retail and service sectors complicates reaching these workers. It would probably not be relevant to find a single strategy for covering all groups.

Table 3.2.8 overleaf outlines the manner in which work environments were assessed in the different areas. Where feasible, workplace hazards were assessed by inspection. Interview (or reported presence of hazards) was done in cases where worksites were not accessible. In the majority of cases (55%) inspection was used to assess hazards, particularly in the production sites. In household surveys and in rural areas particularly workplaces were not accessible so reported hazards were noted.

The profile of work environments is shown in Tables 3.2.9a and b overleaf. Some hazards were extremely infrequently noted or reported and are thus not shown in the tables^{(Endnote 8)}.

Work risks in the home environment (urban)

Mixed workplaces- mixed exposures (rural)

Work in bus terminals - vehicle exhausts and social problems (rural)

Mechanical risks (urban)

Handtools- mechanical and ergonomic risks (urban)

Handtools- mechanical and ergonomic risks (peri-urban)

Poor ventilation and fire risks (urban)

Dust, mechanical and ergonomic risks (peri-urban)

Electrical handtools and poor housekeeping (urban)

Heavy loads (rural)

Poorly organised worksites, exposure of family members, inadequate water and sanitation (rural and urban)

Chemical and mechanical hazards (urban)

Animals (rural)

Ergonomic hazards, long hours of standing (urban)

Noise spillover between sites (urban)

Poor work postures (urban)

Exposure to sun heat and rain, monotonous work (urban)

As evident from the tables, informal sector workplaces do have a wide range of hazards.

The most common hazards identified in 40% or more workplaces are shown in Tables 3.2.10a and 3.2.10b by area and sector.

The most common hazards relate to work organisation, hygiene and ergonomic problems across all sectors. Dealing with this cluster would reduce a significant share of workplace risks identified through inspection or reported by workers.

* Hygiene facilities (toilets, water and food hygiene) are not available across almost all sectors, indicating potential for improvement of these facilities at a collective level. In rural areas the lack of adequate hygiene facilities would be further compounded by the presence of biological hazards from animals.

* In 78% of workplaces work is done outdoors, significantly increased of course by rural farming, but high also (over 60%) in other production sites. Across sectors exposure to sun (and rain) signals a need, where feasible, for adequate protective covering of workplaces, without reducing the benefits of natural ventilation noted later.

* Ergonomic hazards cut across all sectors, particularly long hours of standing, poor work postures (particularly bending and squatting) and handling of heavy loads. Where seats were used they were reported to be uncomfortable. This merits a specific focus on ergonomic hazards in the small scale sector, and the measures that can be used to reduce these problems.

* Poor workplace organisation were noted across all sectors, with associated tripping hazards noted in manufacturing.

* Noise, dust and heat problems in workplaces are noted across all sectors to spill into neighbouring workplaces, affecting workers who are not exposed to these problems through their own work. This indicates a need not only to control these problems at source, but to ensure better workplace siting and pollution control so that the number of exposed workers is limited.

Other more specific hazards were found in different sectors requiring more specific control measures. These are:

Agriculture: pesticides and biological hazards; work with sharp tools

(& rural households)

Manufacturing: work with sharp tools, repetitive movements; chemicals

(& urban production site)

Rural production sites: exposure to traffic fumes (many workers work at bus terminals)

Chemical use was present in 40% of workplaces, indicating a need for information and safety around chemical use in the sector. High risks of solvent use in urban areas and of agrochemical use in rural areas signal a need for specific hazard control and safety interventions and information on these two chemicals types.

The major chemicals used be sector / area were:

Agriculture agrochemicals (pesticides and fertilisers) & rural households:

Manufacturing: paint and varnishes; lubricants and thinners; glues, cleaners & urban production sites: cutting fluids and red oxide and other dyes.

Retail: paint and varnishes; glue, soaps and detergents

Other: perming lotions and relaxers; lubricants and thinners

Rural production sites paraffin, benzine and petrol and red oxide / other dyes

Urban households: paint and varnishes; perming lotions and relaxers; lubricants and thinners.

How were hazards being controlled? The major controls were through natural ventilation in outdoor work (87% of worksites, ranging from 83%-91% across survey groups) or whatever safety features were inherent in the equipment used. Given that the sector uses second hand and often old equipment, the likelihood of in-built safety measures is low, although this would need to be further assessed. Personnel protective equipment (PPE) was poorly provided and often poorly maintained where it was provided. PPE of various types (Safety glasses, apron, welding mask, hard hat, gloves, respirator, safety boots) was provided in 87 worksites or less (less than 5% of the sample), with the lowest PPE provision in rural households (<2% of workers) and the highest in urban production sites (3-18% workers provided for various types of PPE). Given that about 360 workers were doing work where PPE would be needed, and a further 513 involved in farming or gardening, where PPE may be required, this signals a very rate low of protection through PPE.

Enhancing coverage of PPE may be difficult in the outdoor and less well regulated environment of the informal sector. A more appropriate control strategy would appear be to focus on the worksite design and technologies used and identify inbuilt design features that will enhance safety so that these are not left to individual implementation. Notably the need for technology support to the informal sector has been raised in other studies (Mhone 1994)^{(Endnote 9)}.

Incorporating occupational safety into these technologies will enhance safety in the sector.

Workers were asked about their own perceptions of the risks of their work. The responses are shown in Tables 3.2.12a and b. Nearly half of the workers (49%) thought their work was harmful, higher in urban and rural production sites (55% and 62% respectively) and in manufacturing (61%). The risks or areas where respondents reported that work may be dangerous are shown in Table 3.2.12a and b, excluding risks that were reported in less than 5% of cases^{(Endnote 10)}.

The most common perceived risks in both urban (manufacturing) and rural (agriculture) areas were loads, chemicals and dusts and work postures. In rural production sites (and retail sector), infection or poor hygiene were also commonly perceived risks. This compares closely with the risks observed in workplaces discussed earlier. The perceived health effects of these risks in urban (manufacturing) areas were respiratory problems, eye problems and traumatic injury (cuts or fractures), while in rural (agriculture) areas they were respiratory problems, traumatic injury (cuts or fractures) and stomach problems. Respiratory problems relate to the dust and chemical risks described. Injury outcomes are common and appear to derive from a range of risks, particularly cutting tools, falling objects and workplace violence. In rural sites stomach problems relate to poor hygiene. The expected musculoskeletal problems from work loads and postures were noted, although less commonly than expected, including backache (9%) and muscle strain (5%).

The rates of work related injuries are shown in Tables 3.2.13a and b. Reported injury ranged from 103,9 / 1 000 injuries needing medical consultation in rural production sites to 1 774 / 1 000 total work related injuries in urban production sites. Noting that injury rates in the formal sector average 20 / 1 000, the injury rates are much higher than the rates reported in the formal sector. Comparison should ideally be made with injuries with over 24 hours of lost time. In some surveys, such as the UK labour force survey) the comparison is made with injuries requiring medical consultation. Should this more conservative indicator be used injury rates vary from 103,9 / 1 000 in rural production sites to 153,2/ 1 000 in urban production sites.

(a) percentages shown reflect the percent of those reporting work to be harmful, not percent of the total sample

(*) Source NSSA (1995)
(**) Multiplying by a factor of 0,16 based on total formal:informal sector rates

The manufacturing sector has the highest injury rate (206,5 /1 000) and retail sector the lowest (91,4/ 1 000). While these self reported rates are significantly higher than those reported in the formal sector, they also directly correlate with the sectoral distribution in the formal sector (pearsons r, p<0,001). It is thus likely that they are a reasonable reflection of the distribution of injury in the sectors. If the assumption is made that overall injury rates in the formal and informal sectors do not vary significantly, then the overall informal sector rate of 130,6/ 1 000 would need to be adjusted by a factor of 0,16 to relate them to reported formal sector rates.

Using this adjustment factor, while rates are similar in most sectors, the injury rates in informal sector manufacturing are higher than those found in the formal sector.

In fact, the significantly increased rates in the informal sector through surveyed data may indicate equally significant underreporting in formal sector reporting systems. That this is likely has been noted in the past, and could only be verified through similar surveys of the formal sector. It is notable that the households interviewed reported that they were reluctant to stop work on injury or illness, or even to seek medical care, as this would lead to a serious disruption of earnings. Medical care is also costly for workers, particularly in urban areas. It is thus likely that the use of an injury indicator that involves medical treatment provides a relatively rigorous indicator of injury.

 

The high rates of injury suggest that a substantial burden of occupational morbidity is being missed in both the formal and informal sectors.

Tables 3.2.14a and b further explore the pattern of injury in the informal sector. For a selected injury (the most recent and/or most severe), questions were asked to identify the causes, effects and actions of those injuries. Almost two thirds (63%) of these injuries led to lost work time of more than one day (and would thus be reportable under SI 68 1990). The injuries were specifically selected to be the most recent - in fact almost half occurred within the past 3 months, and a further 31% 4-6 months prior to the survey. As noted in other surveys (Reference), recall of injury is greater within a 6 month period, and this would apply in almost three quarters of the reported injuries.

The reported injuries were most commonly crushings and amputations (36%) and other superficial injuries (15%). There was some variation across areas and sectors: Head injuries and sprains, strains and dislocations were common in urban households, sprains, strains and dislocations common in rural households, bruising also common in urban production sites and burns in rural production sites and manufacturing. The causes of injury were mainly mechanical: viz; being struck by falling objects (14%), stepping on or striking objects, caught in or between objects (13%) and other causes (19%). This compares with the frequencies of mechanical causes of injury on the formal sector. Across sectors, falls and overexertion were common in agriculture, while mechanical causes were common across all sectors.

Of the injuries reported, 81% had no permanent effects, in 19% disability was permanent but with ability to carry out the same job, and in 3% permanent disability implied need for a change of job. In no cases were workers no longer able to work. Permanent disability was more frequently reported in rural households (33%) and in agriculture (28%).

In only one case was the injury reported to the workers compensation fund at NSSA, and in no cases was compensation paid. This significant under-reporting of cases to NSSA is not a surprise (and was one motivation for the study). In law (SI 68 1990), all workers with a contract of employment should be covered by workers compensation. This would imply that 297 workers (or 19% of the sample) should have been eligible for workers compensation.

 

It is evident that SI 68 1990 is almost universally breached in the informal sector.

Information was also sought on rates of illness, the results shown in Table 3.2.15a and b overleaf.

The field workers reported that workers were reluctant to answer direct questions on their own ill health at work, preferring to respond to the prior question on the risks in the work and its effects. This may have led to some under-reporting of work related illness. Over attributing common public health problems to work would lead to over-estimation of illness due to work, although fieldworkers also reported that people do not normally associate work with illness, so that this bias is not likely to be significant. A healthy worker effect in production sites would underestimate work related injury and illness, again unlikely to be significant as workers were reluctant to take time off work for occupational injury or illness and because the larger part of the sample was drawn from households. In the survey the number of workers NOT interviewed and the reasons given were recorded. This indicated that 156 workers were not interviewed, distributed by area and cause as shown in Table 3.2.16. Of these 13 (8%) were not interviewed due to illness at the time of the survey, 23% due to their age (too young or too old) and 69% for other reasons, such as being away from home. If these 156 are added to the 1 570 interviewed, the total loss to the sample due to illness is 0,8%. This would only marginally increase the illness rates and only if these illnesses were work related. This was possibly the case in three of the reported illnesses.

(*) Too young (<12 years) or too old (incapable of being interviewed)

Illness rates are similar to injury rates (Illness 117,2/ 1 000 and injury 130,6/ 1 000). Illness rates exceed injury rates in agriculture and are lower than injury rates in manufacturing and retail sectors. Rates of work related illness are higher in rural than urban areas, with peasant households having the highest rates. Aggravation by work is as common as work related causes of illness, with possibility of significant overlap in the two questions. It would seem that respondents had difficulty in separating cause from aggravation.

The most common forms of ill health are respiratory disorders (43%), particularly in urban areas, manufacturing and retail sectors. In rural areas and the agricultural sector musculoskeletal problems and headaches are more common, with eye problems also common in rural production sites. This is highly consistent with the findings of other reports on the informal sector outlined in Section 2.

(*) From respondents indicating others with similar health problems over total workers in workplaces covered

Workers were asked whether other workers had similar illnesses in their workplaces. Using total reported cases over a denominator of total workers, much lower rates of illness are reported (15,6 / 1 000), higher in urban production sites and manufacturing. It is not clear how useful this statistic is except to give an idea of the extent to which workers perceive others to have common health problems. This perception may be stringer in the more collective forms of production, such as urban production sites, hence the higher rates. Lower rates in agriculture may relate to the pattern of musculoskeletal and headache problems in this sector, with workers not likely to know other workers subjective symptoms. Talking about ill health is infrequent and discouraged, as this is regarded as highly personal. It is therefore unlikely that workers would know each others health problems, reflected in the wide variance between self reported and indirectly reported rates.

This bias would probably not apply in the case of mortality, and Table 3.2.18 shows the reported occupational fatality rates. The annual total reported rate of 12,49 deaths / 100 000 workers compares with formal sector annual rates of 25/ 100 000 (NSSA 1995). As the informal sector rate is based on very little data, it would need a larger sample for a more confident estimate to be obtained. It is likely, given the turnover of workers with insecure contracts and high levels of self employment that a healthy worker effect exists in the informal sector statistic. The data thus signals that occupational fatality in the informal sector may be range from half to 100% of the rates in the formal sector.

Endnote 5:
The maximum hours in manufacturing by industrial agreement is 44 hours, such as in the metal sector, and in agriculture is 51 hours, excluding shift workers.

Endnote 6:
The compiled total workforce size shown in Table 3.2.7a for the whole sample is adjusted for the missing data using the average workforce size, so that this information can be used in denominators in the calculation of rates.

Endnote 7:
It is certain that there is some duplication of reporting of the same workplaces from different workers from the same workplace, particularly in the production sites where a census of workers and workplaces was carried out. This duplication would also apply in the numerator data where this total is used (ie: reported numbers affected by illness and reported deaths in the past 5 years). In other rates the denominator used was the sample size, hence avoiding this duplication.

Endnote 8:
Hazards noted in less than 10% of workplaces across all sectors and areas included:

Mechanical: narrow walkways, unguarded machine parts, work with vehicles, hand held machines heavy

Physical: live electric wires, whole body vibration, other extreme heat sources

Chemical: vegetable dusts in the workplace

Endnote 9:
Mhone (1996) notes that "the question of access to technology features only as the fourth ranked problem as perceived by the informal sector participants themselves. However, an objective assessment of some of the key activities that appear to be viable suggests that they lag badly behind in terms of the technology that they use impacting negatively on the quality and general competitiveness of their products. A clearly formulated, targeted and concerted technology selection and dissemination programme is fundamental in order to make the selected informal sector activities achieve their full potential. The issues of finance and the need for other types of support services have need amply documented as cited above. What this points to is the need for a package approach to assisting the selected informal sector activities".

Endnote 10:
These were noise hazards, and burns and electrocution/shock effects

[ Table of contents ]

[