The ILO’s embrace of “decent work for all” illustrates the agency’s interest in incorporating a greater development focus which places the needs of the working poor, and particularly informal workers, more central to its aims [ILO 1999]. Decent work integrates the ILO’s four strategic objectives,2 and can be linked to poverty reduction since increased opportunities for work, increased rights at work, social protection and greater voice in the workplace are associated with improvements in capabilities and well-being [Chen, Vanek and Carr 2004; Saith 2004]. The key point is that it is not just the number of jobs created but the quality of those opportunities that make work decent and which links decent work to poverty reduction [ILO 1999, 2002a and b; Chen, Vanek and Carr 2004].

The ILO has invested in efforts to develop measures of decent work, noting the complexity of the issue and the particular challenges related to the availability of national data for many desired dimensions. Much of the ILO’s work has focused on nationally comparative measures [Bescond et al 2003; Anker et al 2003; Bonnet et al 2003; Fields 2003] but considerable effort also has gone into translating the concept to the micro level through attention to work securities [Anker 2002; Standing 2002; Bonnet et al 2003; Unni and Rani 2003].

The ILO translated decent work to the micro level through the notion of security [Bescond et al 2003; Anker 2002; Standing 2002]. It instigated People’s Security Surveys (PSS) as a means to collect data to understand insecurities from workers’ perspectives. The PSS incorporated seven work-based insecurities along with basic insecurity, which covered insecurities around food, housing, healthcare and the like. Table 1 lists the insecurities and provides details of the indicators for each insecurity as developed in the PSS [Anker 2002] along with the indictors used in Standing’s (2002) more simplified empirical application. The PSS recognises that insecurities vary by employment status so

Economic and Political Weekly May 27, 2006 it posed different questions for different types of workers (salaried, casual wage, piece rate homeworkers and self-employed). This increased the complexity of the instrument but led to measures which better reflect reality.

Overall the seven work-based insecurities capture highly relevant dimensions of work quality and have the potential to incorporate indicators important to decent work for informal workers. The job security dimension is perhaps the least immediately relevant for informal employment. This is particularly the case for indicators representing career niches or professions. Many informal workers end up in jobs or sectors not necessarily by choice, but based on what opportunities are available. Thus, notions of a career or profession may not fit. This is not to say the dimension is unimportant. In the long term the aim is that people will have such choice and will be able to choose a vocation. Job security also captures the need for one’s skills and skill dilution, both relevant in the informal economy. Responses can be disaggregated by relevant social categories to illustrate distributions of work quality.

Anker’s paper details the concepts and methodology associated with the People’s Security Survey. It includes an expansive list of indicators for each security (Table 1). Anker recognises the burden such a list of indicators represents for respondents, so PSSs have tended to be implemented “cafeteria style”, with national research teams choosing modules and question wordings based on the local context. Thus, not all the indicators are used for each dimension of work-based security. This is clear from the indicators Standing used, which also shows that dimensions can be dropped.

Anker makes clear that the PSS must ask relevant questions by employment status; this is apparent in some of the indicators. For example under employment security, asking if one needs a licence is relevant for self-employed workers. The indicators for employment security also include if the worker does subcontracting work which is relevant for piece rate and self-employed workers. More could be done to particularise the work conditions of piece rate and self-employed workers however. For example in Unni and Rani’s PSS application in Gujarat, India they include questions about access to credit, demand sufficiency and scope to expand production underemployment security and incorporate questions about harassment by municipal authorities or local power brokers which fit under work security.3 None of these are included in Anker’s or Standing’s measures. The representation security indicators must not exclude informal workers, particularly

Table 2: Micro Level Decent Work Measures Adopted inTable 2: Micro Level Decent Work Measures Adopted inTable 2: Micro Level Decent Work Measures Adopted inTable 2: Micro Level Decent Work Measures Adopted inTable 2: Micro Level Decent Work Measures Adopted in

the Surat Casethe Surat Casethe Surat Casethe Surat Casethe Surat Case

Decent Work Dimension Measures

Labour market security ALL: Number of days worked and unemployed; regularity of work; hours of work; like to work more hours and/or days; excessive hours and desire for more hours; difficulty finding new work; restrictions on women working

Employment security SAL/CAS: Dismissal notice; type of contract SE/PR: unit/enterprise is registered; have legal recognition; have demand adequacy/access to credit/scope for expansion/regular access to supplies; work for subcontractor/do subcontracting work; have only one supplier or customer; stated dependence on subcontractors ALL: tenure in current activity/workplace; likelihood of losing current work; likeliness of losing work if ill

Job security ALL: Reason for choosing this work; ease of skill transfer; relative skill level

Work security ALL: Perceived health and safety conditions; exposure to dangerous substances or work conditions; work’s effect on general health; access to potable water and toilets; night work; work-HH duty conflict; childcare arrangements SE: harassment over right to space

Skill reproduction security ALL: Source of skill training; length of training received; access to training in current job; perceived usefulness of training

Income security ALL: Earnings from work; variation in earnings; ability to save SAL/CAS: access to non-wage benefits

Representation security ALL: Membership in organisation representing work interests; cooperative membership

Sources: Anker (2002); Standing (2002); Unni and Rani (2003).

Table 1: Micro Level Decent Work Measure Based on the ILO’s People’s Security SurveyTable 1: Micro Level Decent Work Measure Based on the ILO’s People’s Security SurveyTable 1: Micro Level Decent Work Measure Based on the ILO’s People’s Security SurveyTable 1: Micro Level Decent Work Measure Based on the ILO’s People’s Security SurveyTable 1: Micro Level Decent Work Measure Based on the ILO’s People’s Security Survey

Decent Work Dimension Anker’s Detailed Measures Standing’s Measure

Labour market security Employment status, hours of work, multiple work activities, Not included length of experience; days of unemployment; difficulty finding work in case of job loss; notice period; restrictions on women seeking work; perceived likeliness of losing work if pregnant or ill

Employment security Occupation, place and regularity of work; employer characteristics; Perception of security in main work; employment type of contract; do subcontracting work or work for labour status (regular employment,casualearner, outworker); contractor; need for licence; tenure at current work; perceptions size/type of firm; location of work; more than five of satisfaction; expectations of keeping current job/work workers in the establishment

Job security Past advancements or regressions; expectations of need for Improved type of work done in past five years; skills and of advances or regressions in work; perceived expectations of having a good job in 12 months’ time importance of following own profession

Work security Absence from work due to work-related injury, illness, stress; Dangerous equipment or materials; health and safety excessive work hours; control over work; sexual harassment; department at work; opinion of workplace safety; has hazardous work/dangerous equipment; toilet and water available; accident insurance; for self-employed: feels secure use protective clothing; safety department at workplace; opinion in doing business or the work of workplace safety; compensation for injury/insurance for injury; childcare help; absence due to household duties

Skill reproduction security Formal/informal training received; use training/qualifications/ Received formal and/or informal job training; use education in one’s work; opinion on skill adequacy and need for training/qualifications/skills/education in one’s work further training

Income security Income level; fringe benefits; wage arrears; how income received; Respondent thinks income adequately or more than if women keep income; raw materials/equipment provided by adequately covers basic living needs; actual receipt employer; regularity of income; expectations for future income; of income entitled to; regularity of income; opinion on opinion on income adequacy and relative income; ability to save one’s relative income; access to non-wage benefits

Representation security Knowledge and opinions of unions; union in the workplace; belong Belongs to organisation expected to represent to union; knowledge of and/or membership in other worker interests/rights at work; union in workplace; belong to organisations; circumstances for action any other (non-union) organisation representing work

not self-employed or piece rate workers. Both Anker and Standing have taken care to include questions about organisations representing work-based interests, avoiding focusing only on unions.

Another weakness of the PSS measures described by Anker and used by Standing and Unni and Rani is their inattention to terms of incorporation into work, and particularly the issue of dependence among self-employed and piece rate workers. Anker does include a question asking if the respondent works for a labour contractor or does subcontracting work. More could be done to ask self-employed and piece rate workers about the number of sources from which they obtain raw materials and the number of outlets to whom they sell their final products. This will provide evidence of monopoly and/or monopsony conditions, which place the producer in highly dependent relations of production.

The decent work measures used in this paper (Table 2) incorporate all seven work-based insecurities. They are drawn from Anker’s as well as Unni and Rani’s indicators and have been developed with explicit attention to employment status differences and informal employment. They give relatively less attention to job security and add measures of dependent production or employment relations to the employment security dimension.

Surat is a city of 2.4 million residents located in South Gujarat. It is a fast growing urban centre, with decadal growth of 93 per cent from 1981-91 and 60 per cent from 1991-2001 [Das 1994; Census of India 2001]. Surat has a primarily Hindu population and Muslims compose the single largest minority group [Das 1994]. According to census figures, in 1991, 34 per cent of its population had main worker status, which has increased to 38 per cent in 2001.

Surat is particularly known for its diamond cutting, synthetic silk textile and ‘jari’ (gold thread) industries, all of which have links to global and domestic markets. Its vibrant manufacturing sector, which supports strong service and petty trade activity, draws migrants from other districts in Gujarat and from far ranging states. Therefore much of the informal labour force in the city is from outside Surat [Das 1994].

The data used in this paper were collected using a combined household and individual level survey instrument since the study combined two objectives: to understand individual worker insecurities and to estimate the size and economic contribution of the informal economy. The household level survey collected basic data from a household representative about days of work, employment status, characteristics of the work and earnings for all working household members which allowed for the estimates of size and economic contribution. Most of the data used in this paper were drawn from the individual level survey, which involved purposively selecting one working household member from among all working members, based on the distribution of workers and household main income sources across employment status categories, and questioning them about the different forms of insecurities experienced.

The survey was carried out in one to six low-income settlements in each ward in the Surat Municipal Corporation area; some smaller wards were clubbed together. Lanes were selected in each settlement at random and all households residing in the selected lanes were enumerated to document for each household the main source of income (from salary, casual work, self-employment or home-based piece rate work) and the number of working males and females in each employment status category. We interviewed 814 households and individual workers from March to May 2004, 407 males and females each. The selected sample did not fully represent the enumerated households’ employment status distribution due to difficulty in connecting with some workers (particularly casual workers) for interviews even though field staff went to the settlements in early mornings, evenings and Sundays. We over-sampled females because we were interested in understanding gendered experiences of work-based insecurities.4 Because the sample is not random no tests of significance are reported, as we cannot infer to a population.

Men are far more heavily concentrated in casual labour due to the prevalence of casual forms of employment in two of the city’s main industries (textiles and diamond cutting) (Table 3). Women are more evenly distributed across employment status categories than men; women’s greater representation in salary work is driven by the high number of women in domestic service who we classified as salary workers since most received a fixed monthly salary.

The respondents have a long median length of residence in the city, even with the high rate at which families had migrated to Surat (i e, few are natives of the city). The level of migration is also reflected in the range of main languages spoken in the sample households, with high representations of Marathi, Hindi and Oriya, the latter two particularly among the households of male respondents. The workers’ education is relatively low with women faring worse than men. This characteristic implies that many of these workers are ill prepared to obtain higher quality work where formal education is a prerequisite.

Labour Market Security

Labour market security assesses the ability to obtain work. It is measured here by days unemployed, underemployment, regularity of work and the perceived difficulty of finding work. It also incorporates socio-cultural restrictions on working or place of work. Thirty per cent of the sample experienced days of unemployment during the previous 365-day period. Unemployment was concentrated among casual and piece rate workers, and for all but the self-employed a larger share of men than women reported being unemployed at some point in the past 365 days (Table 4). However among those reporting unemployment, women experienced more days of it on average. The average days of unemployment experienced in a year for all those reporting some unemployment was 76; women reported 90 days and men 66.

Obtaining regular work is a problem for 23 per cent of the workers; this is less than those reporting unemployment and illustrates a structural problem in obtaining a fixed supply of work. Irregularity of work is most common among casual and piece rate workers of both sexes (33 per cent of casual and 37 per cent of piece rate), though considerably more men than women report irregularity among casual workers.

Table 3: Percentage Distribution of Sample Workers byTable 3: Percentage Distribution of Sample Workers byTable 3: Percentage Distribution of Sample Workers byTable 3: Percentage Distribution of Sample Workers byTable 3: Percentage Distribution of Sample Workers by

Employment Status, SuratEmployment Status, SuratEmployment Status, SuratEmployment Status, SuratEmployment Status, Surat

Salary Casual Self-Home-based Total employed Piece Rate (n)

Household main

income source 18.2 55.5 20.0 6.3 814 Females 26.5 21.1 21.9 30.5 407 Males 18.2 47.4 25.3 9.1 407

Economic and Political Weekly May 27, 2006

Underemployment is captured through positive responses to questions asking if the respondent would like to work more hours or days than are available. Twenty-nine per cent of the sample would like to work more days and 30 per cent more hours than available. There are differences across activity status, with less than 10 per cent of salaried workers desiring more days, 23 per cent of selfemployed, 35 per cent of casual and 46 per cent of piece rate workers. The only gender difference is among casual workers with more men desiring an increase in work days. The pattern for work hours by activity status and gender is similar (Table 4).

Results for those working excessive hours (over 48 per week) and still interested in working more hours provide further evidence of underemployment and the poor quality of work for casual and piece rate workers particularly. This combination signals extremely poor returns to work. Just under two-thirds of the sample work excessive hours. Across all work activities many fewer women than men report working excessive hours in paid work reflecting their multiple work burdens and significantly longer hours per day in household work (Table 4). Over three quarters of men across work activities, and as high as 90 per cent of male salaried and casual workers, works excessive hours.

The prevalence of excessive work hours does not stop workers from desiring more hours. In fact over one quarter of casual, selfemployed and piece rate workers who work excessive hours would work more hours if available, with the share for piece rate workers climbing to 38 per cent. Only 9 per cent of salaried workers would do so. These differences reflect relative earnings differences between workers in each activity status group (see income security).

Table 4: Labour Market SecurityTable 4: Labour Market SecurityTable 4: Labour Market SecurityTable 4: Labour Market SecurityTable 4: Labour Market Security

Notes: M-male, F-female; Standard deviation is reported in parentheses for interval data. Percentages figures are percentages of total workers in that category (i e, total male salary workers, total female salary workers, total male workers, total female workers).

Table 5: Employment SecurityTable 5: Employment SecurityTable 5: Employment SecurityTable 5: Employment SecurityTable 5: Employment Security

Notes: M-male, F-female; Standard deviation is reported in parentheses for interval data. Percentage figures are percentage of total workers in that category (i e, total male salary workers, total female salary workers, total male workers, total female workers).

self-employed having the lowest share (60 per cent) and casual workers the highest (90 per cent). There are significant gender differences among salary and casual workers; more female salary workers (65 per cent of who are domestic service workers) and more male casual workers felt insecure.

Employment security is also represented by a low share of work years in one’s current occupation spent in the current job. As Table 5 shows, casual workers are the worst, with only about half of their work years in their current occupation spent in the current job. Female casual workers had more employment security than men along this measure. Salary and piece rate workers spent similar time in their present work arrangements while the self-employed, not surprisingly, reported the highest share since their field and job are more intertwined.

Characteristics of the enterprise at which one is employed, or one’s own enterprise contributes to levels of employment security. One important characteristic of an enterprise is whether or not it is registered. Registered units employing casual, salary or piece rate workers should be more accountable for labour regulations while registered self-employed units should have more protection from harassment or exploitation. Sixty-one per cent of the 800 respondents who answered this question stated that the unit at which they worked was not registered; 26 per cent reported its registration through the Factory Act and the balance reported some other form of registration (i e, municipal licence).

Most male salary workers reported being employed by a registered unit while the opposite is true of the females, working mainly as domestic workers. Over 90 per cent of both self-employed and piece rate workers had no registration, or known links to a registered unit in the latter’s case, while only 28 per cent of casual workers reported working for a firm without registration. The high share of casual workers working for firms registered under the Factory Act highlights the use of casual workers in “formal sector” firms. Fewer female casual workers worked in such firms.

Some indicators of employment security are specific to activity status. For salary and casual workers contract type and length of dismissal notice indicate employment security including dependence. Having a contract provides some measure of security and autonomy; unsurprisingly only a minority of casual workers had one, mostly men. Salary workers were in a better position with a quarter having a contract, again mostly men. However, just under half had no contract at all, with the majority of these being women. Male salary workers also come out best in terms of dismissal notice, with a quarter given three months’ notice compared to 9 per cent of women and 4 per cent of male and female casual workers. For casual earners, there is no gender difference in dismissal notice given and the modal category is no notice (40 per cent). These results add to the rather dismal picture of employment security for female salary and all casual workers.

The employment security outcomes of self-employed and piece rate workers are more mixed. Very few, and particularly very few women, have any form of licence or identity card (Table 5). They also face barriers to employment security due to lack of demand or insufficient markets, with over a quarter of both self-employed and piece rate workers reporting this problem. Twice the share of male than female piece rate workers identified this problem.

Other factors Unni and Rani associated with employment security for self-employed and piece rate workers were less problematic amongst this sample. Only 10 per cent of the selfemployed felt constrained by a lack of credit for fixed capital, and half that share felt limited by a lack of credit for working capital. Essentially no piece rate workers felt limited by a lack of credit for either type of capital. The common call for improved access to microcredit seems unfounded in this context and among this sample.

Irregular access to inputs also was not a major problem for either self-employed or piece rate workers. In fact both groups felt highly positive about future potential, with 85 per cent of self-employed and 80 per cent of piece rate workers reporting they saw potential to expand their operation to new markets. Females were somewhat less likely to say this, particularly among piece rate workers. However, a majority of women in both cases still reported optimism. This in some ways contradicts earlier reported general pessimism about the high likelihood of losing one’s current work and may point to the challenge of integrating perception-based questions within a quantitative study.

Experience of dependent production relations also is linked to employment security for self-employed and piece rate workers. Among self-employed and piece rate workers monopoly conditions are common while for piece rate workers monopsony conditions also predominate. Over half of self-employed workers, and more women than men, depend on one supplier for their inputs while almost 90 per cent of both male and female piece rate workers are in the same dependent position in relation to sources of inputs.

All but 3 per cent of piece rate workers obtain their work through a system of subcontracts. Among all piece rate workers, 83 per cent have only one source of orders. Interestingly there is no significant gender difference even though one might expect men to be more mobile and connected to the market than women. The final signal that dependence on contractors is a problem for piece rate workers comes from the result that just over half of males and females self-stated that such dependence was a constraint.

Job Security

Job security relates to having an occupational niche and is measured by the reason why the respondents entered their line of work, the perceived ease with which others can learn their skills and self-classification by relative skill levels. The latter

Table 6: Job SecurityTable 6: Job SecurityTable 6: Job SecurityTable 6: Job SecurityTable 6: Job Security

Notes: M-male, F-female. Percentages are percentage of total workers in that category (i e, total male salary workers, total female salary workers, total male workers, total female workers).

Economic and Political Weekly May 27, 2006 is included here and not under skill reproduction security because those rating themselves as highly skilled would likely have an occupational niche. Skill reproduction security relates more to training received and access to training currently.

Just under 40 per cent of respondents entered their line of work by choice, reflecting their interest or the draw of a traditional family occupation (Table 6). Others entered their line of work because it was what was available or because someone helped them enter. There are significant differences across activity in the share reporting entry due to interest. The largest share is among self-employed workers followed by piece rate, casual and salary workers. The results for salary workers are influenced by the number of women in domestic service who report entry into this work because it was available. Note that a larger share of female than male piece rate workers also report entry in response to availability. These responses reflect the limited types of work open to females in the south Asian context.

All but three respondents felt that others could learn their skills; the difference is in the ease with which they could do so. Fortyfive per cent reported others could easily learn their skills, denoting low job security while 54 per cent reported others could only learn their skills with difficulty. Females in all activities save self-employment were more likely than males to report that others could learn their skills easily, perhaps reflecting some measure of internalisation of perceptions regarding the skill level of typically female work. Male casual and piece rate workers were the least likely to report others could learn their skills with ease. This rather high perception of barriers to skill transfer does not translate into higher relative employment or income security for these workers.

Physical conditions of work are generally acceptable across the workers (Table 7). Those most affected by the lack of these facilities were self-employed workers, and within these largely men as they are likely to work away from home. In terms of general health and safety conditions in the workplace, 55 per cent of all respondents rated them very good or satisfactory, the balance as reasonable or poor. There is some variation across work activity in this; more piece rate workers rated their workplaces (their homes) as very good while more casual and selfemployed workers rated the workplace as reasonable or poor.

Table 7: Work SecurityTable 7: Work SecurityTable 7: Work SecurityTable 7: Work SecurityTable 7: Work Security

Notes: M-male, F-female. Percentages are percentages of total workers in that category (i e, total male salary workers, total female salary workers, total male workers, total female workers).

One final outcome representing work place security applies only to the self-employed. It relates to insecurity in plying their trade, particularly for those doing so on the street whether from a variable or fixed location. We asked if any self-employed respondents (40 per cent who work on the street) experienced harassment from officials, local toughs, etc. Fifteen per cent reported that they did, and among those working on the street, one-third responded positively. Due to women’s greater frequency of engaging in self-employment activities in or near the home, significantly fewer women than men reported such harassment (8 per cent compared to 21 per cent).

Skill Reproduction Security

Skill reproduction security assesses how well workers can gain and maintain skills. We have measured this security through source (formal or informal) and duration of training received, whether individuals can access training in their current job and their perceptions about how useful training is to getting work, obtaining more regular work or earning more. Overall, informal sources of training are the most common (Table 8). More salary workers accessed formal training and within salary workers more males than females did so. More female than male self-employed and piece rate workers had formal training; this likely reflects the targeting of females for NGO and government sponsored microenterprise development programmes.

The prevalence of informal training raises questions of how systematic the training is and how well the knowledge is retained. Such questions are reinforced by evidence of the rather short duration of training received. Reporting medians due to high variation in the data, there is evidence of considerable difference in training duration by whether it was formal or informal; formal training lasted five times longer (measured in days). There are more marginal differences in median days of training by activity; the median across all activities is 30 days, giving rise to concerns about the depth and usefulness of the inputs. Starker differences are apparent within activities by gender. Women are clearly unable to access the same length of training as men; their median days are half or less than men’s, even among female dominated piece rate workers.

Access to training in the current job is limited; just under a quarter of workers reported they could access training at their current workplace. Casual workers were more likely to report access to such training (on the job training from work mates). More male than female salary workers reported access to training at work; this reflects the dominance of domestic workers among the females.

Workers did think training in general is useful. About 60 per cent felt training helps in obtaining work and in earning more while just under half saw its usefulness in gaining more regular work. Salary workers were less inclined to see training as important to getting work or earning more compared to workers in the other

Table 8: Skill Reproduction SecurityTable 8: Skill Reproduction SecurityTable 8: Skill Reproduction SecurityTable 8: Skill Reproduction SecurityTable 8: Skill Reproduction Security

Notes: M-male, F-female Percentages are percentages of total workers in that category (i e, total male salary workers, total female salary workers, total male workers, total female workers).

Table 9: Income securityTable 9: Income securityTable 9: Income securityTable 9: Income securityTable 9: Income security

Notes: M-male, F-female; Standard deviations are in parentheses. Percentages are percentages of total workers in that category (i e, total male salary workers, total female salary workers, total male workers, total female workers).

Economic and Political Weekly May 27, 2006 activity categories. More casual workers than any others perceived training as important to these two outcomes and more males than females expressed this opinion. Similarly more male than female self-employed workers thought training could lead to more work or better earnings.

All in all training is important to the workers, though few accessed formal training and the duration of training received was not long, particularly for female workers. Skill building and skill appropriateness are areas that require further inputs to improve workers’ abilities to integrate themselves securely into changing labour markets.

The evidence of the above insecurities has implications for income levels and stability. To standardise for differences in days and hours of work, hourly earnings were calculated based on reported pay levels and days and hours of work. Across activity type, piece rate workers earn less per hour than the other workers;5 their median hourly earnings are only Rs 4, less than half that of the others (Table 9). One interesting item to note is that while salary workers seem to have better work quality along other dimensions of decent work, they do not earn more per hour than casual or self-employed workers. Some of this may be due to the effect of women’s lower pay on the mean value in this category.

Disaggregating by gender shows different rankings for male and female salary workers, as is the case with many of the other dimensions of decent work. Examining median hourly earnings by activity category and sex shows that among men, salary workers earn the most, followed by self-employed and piece rate workers and finally casual workers. Among women the order is self-employed followed by casual earners, salary workers and piece rate workers. The gender disadvantage in earnings is quite clear. Controlling for hours worked, across all activities women earn considerably less per hour than men; the largest differences are among piece rate and salary workers where the difference in median hourly earnings is more than one rupee. Occupational segregation plays a role with female salary workers being concentrated in low paying domestic work and many female piece rate workers involved in subcontracted outwork creating products out of jari.

Examining daily earnings against the range of official minimum wages in Gujarat6 provides further evidence of income insecurity for female workers. Male workers in all activity categories have median daily earnings well above the maximum minimum wage in Gujarat (Rs 85.6). In fact only 5 per cent of all male workers earn less than the minimum value of the minimum wage per day (Rs 46.8) and 23 per cent earn less than the maximum value. Compare this to females among whom both salary and piece rate workers have median daily earnings less than the minimum value while the others are not far above it. Overall 62 per cent of women earn less than the minimum value of the minimum wage and almost 85 per cent earn less than the maximum value.

Income stability is another dimension of income security. Examining income variability shows that only salary workers of both sexes have any level of income stability. This is measured by calculating the range of income earned (maximum less minimum earnings for the last year) as a share of the maximum. Salary workers experience a variation of only approximately 5 per cent on average, compared to 32 to 43 per cent for the others (Table 9).

particularly require attention as most earned far less than published minimum wages in the state. Two particular subgroups of women workers on which to focus are domestic workers and piece rate workers. For the latter, the ILO convention on homework is in place but needs to be enforced. Enforcing minimum wages for domestic workers is challenging due to the dispersion of workers and employers within private homes. Support to build and sustain organisations of domestic workers is needed to increase their voice and bargaining position. The need for worker organisations extends across work types and gender and is one way to increase autonomy and power among informal workers.

Underemployment is another problem crossing work activity though affecting salary workers less. How to address underemployment varies across activities; for piece rate and selfemployed workers addressing underemployment requires more numerous and autonomous links with suppliers and contractors while for casual workers and female salary workers it is linked to obtaining contracted work for specified periods, with clearly stated dismissal criteria. Addressing both pay levels and low and irregular days of work may reduce male workers’ need to work excessive hours to earn sufficient income. Enforcing regulations regarding work hours and overtime would also influence this outcome for salary and casual workers.

Training can also improve pay levels and access to more and regular work as higher skills can increase competitiveness in the labour market. Workers perceived a clear pay-off to skill training as many connected better skills to more work, more regular work and better pay. The challenge is in ensuring the quality and demand responsiveness of the training.

Other means of addressing decent work deficits are activity specific. Linked to underemployment, casual and female salary workers need some formalisation of their work through access to written contracts and fair dismissal terms; this should ideally make them eligible for basic non-wage benefits, particularly related to medical care, maternity benefits and paid holidays. For self-employed and piece rate workers, more autonomous links to markets are needed to reduce the potential for exploitation; this may require improved access to information about markets and market linkages. State-sponsored licensing procedures can also improve the legal status of these workers and qualify them for any state benefit schemes. Finally, for male self-employed workers particularly, improved provision of basic services is needed. This group had the worst access to potable water and toilets while at work. Urban planning processes have to do more to plan for the informal economy, reflecting its contribution to the vitality of the city.

m

Email: jeemolunni@yahoo.co.in

NotesNotesNotesNotesNotes

coining in the 1970s through to today see the following: Hart 1973; Tokman

1978; Sethuraman 1976; Chen, Vanek and Carr 2004; Moser 1978; Portes,

informal workers. They develop a concept of economic security and use

work deficits, such as caste and age. However, due to space constraints

we focus here on differences by gender.

ReferencesReferencesReferencesReferencesReferences