Estimating Workforce and Understanding Its Characteristics:

An important objective of conducting TUS in most developing countries is to get improved estimates of workforce and to u nderstand its characteristics,8 as these countries consider the e stablished labour force surveys inadequate to net SNA work a ccurately and believe that TUS are capable of giving b etter results.

The concept of informal employment has shifted from the concept of informal sector in the 15th International Conference on Labour Statistics (ICLS) to the concept of informal employment in the 17th ICLS. The informal sector is an enterprise-based concept, while informal employment comprises the total number of informal workers working in formal and informal enterprises, including household enterprises. Informal employment is highly heterogeneous in nature and is difficult to capture through the established surveys. This is because (1) it is not always easy to distinguish between household work and SNA work, with the result that women’s SNA work is frequently underestimated, (2) sociocultural biases on the part of women as well as investigators frequently lead them to report their SNA work as household work, and (3) informal work, which is frequently temporary, seasonal or of short duration; is scattered and sporadic; and is frequently irregular and mobile, is not likely to be netted through the established surveys. TUS, which collect comprehensive data on how people spend their time, tend to have less biases and better netting of SNA work of men, and, particularly, women (ibid).

In addition, TUS can capture information on some specific characteristics of informal employment, such as multiple and scattered nature of work, different work-time arrangements (WTAs), etc. Many informal workers perform multiple jobs in a scattered manner, as they do not get one full-time job. Since these multiple jobs restrict their specialisation, limit their upward m obility and reduce their chances in the labour market, it is important to capture these characteristics of work. The established surveys can capture one or two subsidiary occupations, but are unable to capture multiple and scattered jobs that men and women do. Again, informal work usually has flexible WTAs, the details of which are not captured adequately by established surveys. In addition, no established labour force survey collects data on subsistence workers. Though this work is included in the SNA, it is still outside the purview of these surveys. TUS, however, are able to capture all these details with the help of a proper classification of work and suitable context variable. In short, TUS (a) give improved estimates of workforce, (b) throw useful light on the characteristics of the workforce, and (c) provide estimates of s ubsistence work and workforce (ibid).

It is to be noted that in the case of India, the work participation rates (WPRs) based on the TUS 1998-99 (as calculated by the C entral Statistical Organisation (CSO)) did provide higher e stimates of the workforce than the estimates provided either by the National Sample Survey Organisation (NSSO) (19992000) or the Census of Population (2001) (Saha 2003), (Hirway and Jose 2010).

A related advantage of TUS data is for understanding and estimating child labour and children’s work. Apart from giving improved estimates of child labour, TUS throw light on intensity of their work (in terms of the hours put in) and multiplicity of their jobs. TUS also throw useful light on “nowhere children”, who are neither in the school nor in the labour market. As studies pointed out, a significant portion of these “nowhere” children (particularly girls) are engaged in unpaid household work, and many of them while away their time doing nothing, as schools are not attractive to them or they did miserably in schools (Hirway and Thakar 2006; Rastogi 2008). It is clear that these results are very useful while designing programmes for universal elementary and secondary education.

It needs to be added, however, that the TUS are not a substitute for a labour force survey. It complements this survey.

Measuring Socio-economic Changes: Time is a precious fundamental resource, and its use reflects the progress, achievement and well-being of individuals, families and societies (Ironmonger 2008). Since all social and economic systems involve the use of time, these data have the potential for explaining the dynamic functioning of most of these systems. Consequently, like all other data, time use statistics can be used in tracking (observing and monitoring past performance), researching (establishing connections and understanding relationships), forecasting (predicting future performance of the system) and policymaking (changing future outcomes through policy alternatives and movements of control instruments). These data are, therefore, useful in research, modelling, forecasting and policy simulation. Creating time series data on time use has helped developed countries in measuring and monitoring changes. The data can thus be used in development planning, social policy and research. The data can provide useful insights in the areas of education, health, poverty and overall human well-being, destitute, etc.

In India, time use data can be used as indicators of well-being and poverty. The recent literature on time poverty (Charmes 2004; Kes and Swaminathan 2006, Blackden and Woden 2006) has argued that the poor including poor women in developing countries frequently suffer from time poverty, i e, time stress, and unless it is addressed through right kind of interventions, one cannot address their poverty adequately. It is also argued that when one examines the time use of the poor at the bottom, one observes several severe constraints experienced by them. This is because time is a major asset of the poor, who have no/low assets otherwise, and their poverty is also reflected in their time use. These constraints have important policy implications for poverty reduction s trategies (Hirway 2009).

Time Use and Concerns Related to Human Development:

A nother contribution of TUS data could be that they can reveal the concerns and constraints related to human development. To start with, the data can provide new evidence on informal i nstruction of children in the home and can assess the relative importance of household attributes and local educational quality for school attendance as well as human capital investment time. Using the pilot TUS data for the six states, along with the data on school quality and availability from the Seventh All-India School Education Survey for the same six states, Sripad Motiram and Lars Osberg have estimated the total time invested in human capital acquisition by boys and girls in the rural areas. They have shown that the human capital investment time of children is i nfluenced by their scheduled caste status, parental education and quality of schooling (Motiram and Osberg 2008). A major conclusion of the analysis is that the inequality in human capital investment time in rural India can be explained, to a considerable extent, by the poor quality of schooling.

Valuation of Unpaid Work: Unpaid work is of two types, unpaid SNA work and unpaid non-SNA work. The former, which includes informal work (work performed by unpaid family workers in household enterprises) and subsistence work, is covered in national income accounts – at least at the conceptual level; while the latter, which includes household work, care of children, old, sick, etc, in the household and voluntary community work, is excluded from the purview of these accounts. The UN-SNA 1993 as well as the Platform for Action (PFA) declared at the Beijing World Conference of Women have recommended valuation of unpaid non-SNA work into satellite accounts to measure contribution of this work to the total gross domestic product (GDP) in the economy.9 Valuation of unpaid work in money terms, however, is opposed by some scholars on the grounds that (1) it will be like clubbing apples and oranges, as both the works are performed under different situations (the SNA work is performed under market environment, and the non-SNA work under the household environment, where time is elastic – work can linger on for a long time – and where there is no pressure for earning profits), (2) it will i nvolve several unrealistic assumptions, such as the quality of services remain the same in all households (for example, the meals cooked, the care provided), or the quality of the product is as good or as bad as the similar product in the market, and (3) it will involve serious problems about the prices to be used and methods to be applied in valuation, as there are no clear solutions to these. In spite of these problems, the valuation can be justified on the strong grounds that without the valuation, women’s contribution to the total well-being of the economy will not be visible. Since the invisibility of women’s unpaid non-SNA work is at the root of the unequal power relationship between men and women within the household and within the economy, it is important to make women’s unpaid work visible in the same units in which paid work is measured. Valuation gives visibility to

december 5, 2009 vol xliv no 49

w omen’s unpaid work and provides a basis to them to claim their due share in the state exchequer. It will also help in engenderment of national budgets by justifying larger allocations to women’s development and empowerment.

Valuation of unpaid work can be done either by the input method, i e, valuing the labour input that goes into the work or by the output method, i e, valuing the output generated by the work.10 However, the input method values only one input, i e, labour, and ignores the rest of the factors. Experts, therefore, have recommended a value added approach, which is frequently known as the output approach in the literature. This approach allows valuation in a satellite accounting framework, under which each unpaid activity is transformed/translated into a measurable output/unit that is prevalent in the market and relevant for household output. Its gross and net value will be calculated by taking care of all taxes and depreciation of capital, etc. The imputed estimates thus perpetrated may be shown in satellite accounts, which are designed to cover the total labour and capital (including land) in non-SNA production. This method c overs total labour and capital and fits very well in the national income accounting framework (Sharma and Kumar 2008).

Time Use Data for Macro Policymaking: Since TUS provide data on the total economy constituting paid and unpaid activities, they can be used in macroeconomic modelling. Antonopoulos and Fontana (2008) have used a gender aware Social A ccounting Matrix (SAM) framework to understand the role of unpaid work in household economies. The authors have presented the framework of a gender aware SAM model that helps in the analysis of what happens to the whole economic system when some unpaid household activities become visible and are paid through public works. Hirway et al (2008) have also studied the multiplier impact of substitution of unpaid work of women through the National Rural Employment Guarantee Act (NREGA) works on a village economy through simulation exercise in the framework of a village SAM. TUS data are extremely important in these models as they reveal the gender division of unpaid labour contributions. They also make evident sectoral “employment deficits”, especially in public services delivery. A SAM that makes the “invisible” part of the economy evident is expected to yield findings that can better inform policy and also improve the efficiency of public goods and services; but also perhaps, more importantly, it can be useful in introducing new criteria according to which such programmes ought to be designed and evaluated. Reducing the unfair time-tax burdens placed on some people ought to b ecome part of the policy dialogue. Documenting such burdens by collecting information on time use is a big first step towards that end.

In this context, it is important to note that human activities can be divided into three categories, namely, SNA activities, non-SNA activities and personal services (which are by definition nondele gable). Most national statistical systems, including the Indian system, collect data only on the first category of activities and formulate macroeconomic policies based on this database. This is not valid firstly because there is no water-tight compartment between SNA and non-SNA activities (also among all the three

Economic & Political Weekly

EPW

c ategories of activities), as activities move from SNA to non-SNA and vice versa, depending on specific economic and socio-cultural situations. For example, non-SNA activities like cooking, washing, cleaning, etc, move to the market when household income/ employment rises or when there is a boom in the economy, and move out of the market in the event of loss of income/employment. That is, SNA and non-SNA activities are organically interlinked and together, they constitute the total economy. Macroeconomic policies have an impact on both, though differently, and this makes it necessary to understand the impact of macro policies on non-SNA work to monitor and design them. Also, non-SNA work has an impact on the macroeconomy in multiple ways: Household care, by providing care to workers and by taking care of children’s development, contributes to human capital formation by replenishing the labour force. Consequently, to include this work when it is in the market and exclude it when it is outside the market does not make any sense.

Exclusion of unpaid services from the general production boundary of the UN-SNA on the grounds that (1) these services have limited repercussions on the rest of the economy, (2) it is d ifficult to impute monetary value of these services, and (3) their i nclusion will have adverse effects in the usefulness of the a ccounts for macroeconomic analysis and policy purposes (UN 1993). These arguments do not seem to be acceptable, first b ecause these services do not have limited repercussions on the economy, because they contribute not only to the total welfare/ well-being of people, but also to human capital formation that promotes economic growth; second, though there are problems with respect to the prices for their valuation, these can be r esolved; and third, if their rightful inclusion in the macroparameters (like GDP, workforce) challenges the utility of macroeconomic analysis, there is a need to improve analytical tools of macroeconomics rather than exclude this work from m acroeconomics!

The theories of economic growth that treated unpaid non-SNA work as “unproductive work”, “leisure labour” or “consumption” and excluded it from the process of economic growth and from the purview of macroeconomics, are no more acceptable. It is now accepted that households are not only consumers, but are also producers. What is important to underscore here is that non-SNA work is a part of the total economy, and that SNA and non-SNA together form the macroeconomy. The question as to how to integrate unpaid work into macroeconomics is therefore an important area of inquiry.

In short, there is a strong case for expanding the statistical paradigm (i e, the way the national economic system looks at the socio-economic life of people) to include unpaid non-SNA work into it. TUS will be the main tool for the purpose.

It needs to be noted, however, that the inclusion unpaid non-SNA work within the general production boundary and recommendation of compiling satellite accounts of unpaid domestic and voluntary services is not the end of the story. Though this is definitely a progressive step in terms of providing visibility to unpaid non-SNA work, it is not adequate because (1) non-SNA work is not yet considered a part of the macroeconomy (it is still put outside the main accounts – satellite accounts), and (2) the GDP still remains truncated, it provides estimates for a part of the economy only. It needs to be underlined that GDP is the measure of the value of all final goods and services produced in a country during a year and not the market value of only those goods and services that are bought and sold in the market. It is necessary, therefore, to mainstream unpaid work into national income accounts ultimately (Monsod 2008).

2.1 Pilot Survey

The pilot TUS conducted in 1998-99 avoided many weaknesses frequently observed in TUS in developing countries. First, the sampling procedure was systematic and the sample size was valid, with the result that the sampling error was found to be negligible (Pandey 2000). Second, the time sample was drawn systematically and was representative in terms of selection of days (of reference week) and the seasonal rounds of the survey. The results, therefore, could provide representative estimates of the time use in the country (in the six states). Third, instead of selecting one or two members from each household (as some d eveloping countries have done), the Indian survey selected all members (6 + years) of the sample households. This allowed an intra-household analysis of the time use. And fourth, in order to avoid any confusion on the part of investigators, while conducting this new and relatively difficult survey, a detailed instruction manual was prepared for field investigators, followed by i ntensive training at the all-India level and at the state levels in local languages. Strong monitoring by members of state and central committee was also organised to see that the survey proceeded well.

Data Collection: Collection of time use data posed problems first because this survey was different from the conventional surveys, and second, because several socio-economic characteristics of the Indian society posed problems. For example, expecting all respondents to fill in time diaries was not valid as the literacy level is low in the country. Or, getting accurate estimates of the time use was not always feasible as many people in remote areas do not use time pieces. The committee decided to take several steps in this regard.

* Getting the right response on the time use from respondents was a major problem in India because of the low levels of literacy in the country. It was decided therefore to use 24-hour time diaries to be filled in by investigators based on the one-day-recall of

60 respondents.12 That is, instead of depending on self-reported time diaries, a decision was taken to employ investigators for collecting information on time use.

An important aspect of the Indian TUS was that the quality of the data depended heavily on the investigator, as the role of i nvestigators was critical in the survey. It was necessary that

(1) the interviewer helped but did not determine the time use, and (2) he/she had established good rapport with respondents. Preparation of a detailed instruction manual and intensive training of investigators was therefore important, and both were organised for the survey.

Classification of Activities: The technical committee designed a suitable time use activity classification using the available classifications on the one hand, and the needs of the country on the other. The available time use activity classifications at the time of conducting the pilot survey included (1) the classifications used in developed countries, (2) classification proposed by the International Labour Organisation (ILO) in the expert group meeting at United Nations Statistics Division (UNSD) New York (1997),13 and (3) the trial classification developed by the expert group meeting. The TAC developed a three-digit classification for the TUS, where the first digit referred to the main categories, the second digit referred to sub-categories and the third digit referred to time use activities.

The full three-digit classification was pre-tested to see whether it included all activities, and a stock-taking exercise was done at the end of the first round to see what changes might be necessary. Additions at this point included resting due to sickness, and forced leisure. The latter was added to distinguish between

december 5, 2009 vol xliv no 49

l eisure time spent through choice and leisure “forced” through lack of available work opportunities.

Data Collection for Valuation: Keeping in mind the needs for valuation, some additional data were collected during the survey. These data were collected at the district level and these included data on wage rates of different categories of unskilled workers in rural and urban areas.

Context Variables: Context variables provide the context of time use activities. The pilot survey used two context variables:

(1) whether the activity was paid or unpaid, and (2) whether the activity was carried out within home or outside home.

In spite of these steps, there are some lessons to be learnt from the experience of the pilot TUS (Hirway 2000b and 2003). We discuss these in brief in the following paragraphs:

2.2 Lessons Learnt from Pilot TUS

Objectives behind Conducting TUS: The first important lesson learnt from the pilot survey is to determine why we want to conduct TUS periodically. The objective of a TUS cannot be a “collection of information on how people spend their time”. Though this was not the objective of the pilot survey,14 there is a need to identify major objectives of future TUS in consultation with policymakers and data users. The objectives must spell out how the TUS data will help in understanding some of the critical concerns of the national economy and society. That is, there is a need to show that the TUS has the potential to throw useful light on some of the important concerns of the Indian economy and that it can be useful in policymaking and monitoring in some of the important areas of the economy. Such clarity will ensure commitment of the government, including its different wings, to the TUS on the one hand, and will have important implications for the content, concepts and methodology of the TUS on the other hand.

Analysis of the data of the pilot TUS done in a few official s tudies (Nath 2003; Saha 2003) and a large number of studies done by other scholars (Charmes 2005; Maitra and Rai 2006; Hirway and Thakar 2006; Chakraborty 2008; Motiram and O sberg 2008; Hirway 2008; Hirway and Jose 2008; Rastogi 2008 and others) have clearly shown that TUS data have a good potential to provide some good insights in some of the major concerns of our economy. There is now a need to go further and articulate these uses in the shape of focused objectives of the next TUS surveys (the proposed objectives are discussed in the next section).

Along with identifying critical objectives of the future TUS, there is a need to involve all stakeholders in determining the o bjectives of the TUS through a proper process of consultation. Such a consultation should involve representatives from all relevant government ministries and departments as well as outside e xperts and potential data users. The government should be aware of the utility of the TUS data and should be committed to use the data.

The decisions about the concepts and methods including time use activity classification will depend on the objectives of the s urvey (Bittman 2008). If we accept that periodical TUS in India will address some of the important national concerns, TUS in

Economic & Political Weekly

EPW

I ndia will have to have certain features. To start with, it has to be an independent standalone TUS and it should be conducted p eriodically to track and monitor these concerns.

Independent or Modular TUS: Whether the TUS should be independent standalone surveys or modular surveys will be an important question. Though several developing countries have conducted modular TUS, the overall experience suggests that independent surveys are preferable. Modular TUS or adding a module of TUS in an ongoing survey has several advantages like it is easy to institutionalise TUS in the data system, and it is less costly than conducting a full-fledged TUS. However, there are several disadvantages of this approach. First, under this approach stylised questions are asked to respondents on how much time they spent on the listed activities during the reference day/days/week. As is observed in the time use literature, many important activities are usually left out of these lists, with the result that limited information is collected; the response to stylised questions frequently d epends on the interest of the respondent (the respondent tends to over report the time on activities of interest and under reports the time spent on uninteresting activities); comprehensive information on time use is not collected as the focus is only on the listed activities (which are usually related to the main theme of the concerned national survey); and recall is usually a problem when the reference period is more than one day or a week. Independent time use surveys, on the other hand, collect comprehensive information on the time use of the reference population without missing out any details. Though an independent s tandalone TUS is relatively expensive, its start-up costs are high, once institutionalised, it may prove to be less costly.

This does not rule out small-scale TUS when necessary. TUS may be conducted for specific socio-economic groups (for example, scheduled tribe population), specific regions (for lagging states), for specific sectors (for example, agriculture or some home-based work) or on specific issues. Such surveys can be comprehensive surveys or modular surveys on a small scale, focusing on specific areas.

2.3 Lessons in TUS Methodology

Some major lessons to be learnt from the pilot TUS in the area of TUS methodology are as follows:

Background Schedule: An important lesson to be learnt from the pilot survey is to improve the designing of the background schedule, keeping in mind the objectives of the TUS. There is a need to establish close links between the objectives and the potential uses of the TUS on the one hand, and the kind of information to be collected in the background schedule on the other. The recommended additions to the earlier background schedule will be (a) assets of the household, (b) equipment/technology used for carrying out unpaid non-SNA work like cooking, washing, cleaning, etc, (c) wages received for similar activities in the local labour market, and (d) details of the presence of old, disabled or sick person in the household.

The background schedule should also include information on household capital stock (domestic appliances, etc), consumption of market services that substitute for household labour (use of maid, childcare centres, nursing care, etc), detailed childcare arrangements, eating out or buying meal from outside, etc. This information is useful in the analysis of the changing boundary between home and market, changes in the domestic decision of labour and outcomes of many of these processes of children. C ollection of data on informal care is very important, as it reveals a lot how society organises care of the old, sick, disabled and others. The background schedule should collect data on those who need this care and how it is organised. How the total care work is divided between household, community, government and market is an important issue of the study.

Treatment of Simultaneous Activities: It is feared that the pilot survey could not collect complete data on simultaneous activities performed by the population. This is because investigators asked a question on simultaneous activities at the end of investigation and during the course of investigation. However, it will be useful if this question is added for each activity reported by respondents, like a context variable. This will help in getting improved estimates of simultaneous activities in the country.

Again, the analysis of the TUS data on simultaneous activities divided the total time spent on these activities by the number of activities carried out simultaneously. This is also not satisfactory as it was assumed that all the activities are equally important. It will help if simultaneous activities are analysed separately to understand their nature and character.

Context Variables: This is another area where there is a need to learn a lesson from the pilot survey. The pilot survey used two context variables: whether the activity was paid or unpaid and whether the activity was carried out inside or outside home. Though these variables were found useful, they were inadequate to understand some of the important contexts of time use activities. It is also felt that we did not tap the full potential of context variables. It will be useful to remember (1) different variables can be used for different types of activities (for example, specific variables can be used for identifying whether SNA activity is an informal activity), (2) multiple context variables can be used to capture details of some of the activities, and (3) new context v ariables can be added for some specific activities. In short, there is a good scope for improvement here to tap the potential of c ontext variables.

Classification of Activities: One more lesson to be learnt from the pilot survey was relating to classification of time use activities. The classification used in the pilot survey was not found fully compatible with the established labour force classification; with the result it was not possible to identify the activities where the TUS estimates of workforce were higher than that of the e stablished surveys. The department of statistics had set up an expert committee (under the chairmanship of K C Seal), which has designed a new classification for time use activities for the country. The new classification has been made compatible with the established classifications and is suitable to meet India’s specific needs. It has some important additions like (a) time spent on u sing different public services, (b) waiting for accessing services, (c) mode of travel, etc. The revised classification is ready for use. The report of the Seal Committee should be approved and should be used for conducting future time use surveys in India.

2.4 Lessons in Data Collection and Analysis

It needs to be recognised that time use data as collected in developing countries are not likely to be accurate to the decimal points. As pointed out by Seal, estimates-based time use survey are to be treated as “plausible estimates” and not as point estimates, particularly in developing countries. (This statement can be applied to many other data sets in developing countries.) There is, therefore,

december 5, 2009 vol xliv no 49

a good scope for improving the quality of TUS data. Since an important observation from the first survey was that the quality of the TUS data depended heavily on the interviewers/investigators, it is clear that more attention should be paid to hiring and training of investigators and to monitoring their work to ensure improved quality of data. The future TUS can work out a strategy in this direction.

Improved quality can also be promoted through some of the following steps:

Analysis of Time Use Statistics: There is a need to design analysis of time use statistics in consultation with policymakers and data users. It is important that the analysis throws light on some of the critical concerns of the Indian economy. Analysis of the time spent on human development (education, health-related a ctivities, etc), analysis of simultaneous activities, seasonal variations in time use, compilation of satellite accounts, estimating informal and subsistence workforce, etc, are some of the proposed areas of analysis. In addition, analysis of the data at disaggregated levels – at the state level and for different social groups – will also be very useful.

Dissemination of TUS Data: There was a good amount of delay in the dissemination of the data of the pilot TUS. One of the reasons for this was perhaps the fact that this was a new survey and the quality of the data was not tested. We, therefore, recommend that the data should be tested, and like in the case of other datasets, released at the earliest.

3.1 Objectives of TUS

The first task is to determine the objectives of the TUS in India. The main objectives of future TUS are described below:

Measuring All Forms of Work: The first major objective of TUS will be to measure all forms of work that is performed in the economy. This will include SNA work and non-SNA work. Time use data are

Economic & Political Weekly

EPW

likely to provide improved estimates of workforce along with a dditional insights into the nature and characteristics of informal and subsistence work and workers. Data on non-SNA work will measure non-SNA (unpaid) work performed by men and mainly women in the economy, and its contribution to total well-being in the economy.

A relative objective will be to understand intra-household sharing of SNA and non-SNA work and study unequal power r elations between men and women within the household as well as within the economy. The data will also throw light on sharing of paid and unpaid work by the poor and non-poor and present a meaningful comparative analysis of these households.

Understanding Issues Related to Poverty, Human Development and Exclusion: TUS data will help in understanding time poverty as well as time use of the poor that reflects their constraints and problems. The data will also throw light on livelihood strategies of informal enterprises, time use of the unemployed and economic contribution of non-SNA unpaid work to the economy. These will have important implications for designing interventions for poverty reduction.

TUS data will also help in understanding (1) low enrolment of children in schools and their high dropout rates, (2) constraints of girl children in attending school, (3) time use of teachers, and

(4) the time spent by children to reach school. Improved understanding of “nowhere” children, who neither go to school nor are in the l abour market, will help in designing interventions for promoting universal education for children. Similarly, the data on time spent on health will provide estimates of the incidence of sickness, time lost by sick persons as well as by those attending to them, time spent on travelling to health services, on waiting etc. All these data will help in understanding the status of health services as well as how to improve access of people to these services.

Understanding the Nature and Extent of Gender Inequality:

TUS will be an important objective in the context of understanding the nature and extent of gender inequality. By opening up the household by providing data on sharing of the different categories of paid and unpaid work by men and women, TUS will provide important insights into the dynamics of gender inequalities. This will have important implications for designing policies for promoting gender equality.

Valuation of Unpaid (non-SNA) Work: In spite of the conceptual and methodological limitations, valuation of unpaid work is important, as it values unpaid work in the same unit in which SNA work is valued. Valuation of work (a) estimates contribution of unpaid work (mainly of women) to total well-being and total GDP in the economy, (b) provides the value of unpaid community services to the national well-being, and (c) forms a basis for the claims by unpaid workers on the state exchequer.

Designing and Monitoring Macroeconomic Policies: Another major concern is related to designing and monitoring macroeconomic policies. The total economy consists of the market economy or the paid economy and the unpaid economy, and macro policies impact on paid and unpaid economies. It is essential therefore to collect data on both the components of the economy. Policies formulated on the basis of the database of the market economy will be far from valid for the total economy. For example, if a reduction in public expenditure or privatisation increases the cost of a service (such as health, education), it is bound to increase the non-SNA unpaid work of the affected households, and particularly of women members of the households. If trade liberalisation allows cheap imports, leading to unemployment of some workers, it is bound to increase the burden of unpaid work of the members of the affected households. Unless this adverse impact on unpaid workers, mainly women is addressed, there will be loss of wellbeing as well as depletion of human capital of women. To collect comprehensive data on total work is therefore essential.

It needs to be underlined here that this use of TUS data is particularly relevant today when the world over financial crisis is likely to have impacted adversely on women’s welfare and their productive capacities.

In short, time use statistics provide a sound basis for formulation and monitoring of major national economic policies.

3.2 Conceptual and Methodological Issues

We have already discussed the concepts and methodologies used in the pilot TUS and the lessons to be drawn from the pilot survey. These primarily referred to (1) the type of TUS (independent stand alone or modular survey, (2) survey design that includes sampling including time sampling, the type of survey instruments (time diaries or stylised analogues) etc, (3) designing of the background schedule, (4) the mode of data collection and steps to improve it, (5) time use activity classification and context variables, (6) analysis of TUS data, etc. We do not want to repeat these in this section.

We would, however, like to underline the fact that there are several unresolved issues with respect to harmonisation of methodology including time use activity classification at the global level. Careful reviews of TUS in different countries, and particularly in developing countries have shown that (1) there are several quality problems of time use data and (2) due to the

Notes

1 This “Non-working Time” is no more considered as non-working by feminist economists.

2 A large number of developing countries have now conducted their first national TUS, most of them since the 1990s. These countries include Argentina, Mexico, Chile, Brazil, Nicaragua, etc (South America); Benin, Guinea, Mali, Malavi, Madagascar, Mauritious, Morocco, South Africa, Kenya, Chad, etc (Africa); and India, Nepal, Bangladesh, Thailand, Mongolia, Sri Lanka, Lao PDR and South Korea (Asia). Time use surveys are spreading fast in the developing world.

3 The survey showed that (1) women’s WPR was higher than that given by the NSSO surveys, and in some regions women’s WPR was higher than that of men, particularly in poor households, (2) though sometimes social customs preclude women from undertaking outside work, their economic (SNA) contribution was not less, (3) children, and mainly the girl child, take part in SNA work in poorer households, and (4) addition of a ssets like a buffalo adds to the total hours of work. Devaki Jain has noted that the findings of the study did make an impact in terms of giving visibility to women’s unpaid SNA and non-SNA work.

64 d ifferences in the methodologies used, there is limited cross country comparability of time use data. A lot of work needs to be done in this area, and India can play an important role here by c ontributing to harmonisation of the concepts and methods ( Hirway 2008; Budlender 2008; Esquivel and Folbre 2008).

Priorities and Cost-related Issues: It is frequently argued that TUS is not a high priority survey for India, as there are many other areas which demand higher priority. It is also argued that TUS is a costly survey and the country cannot afford mainstreaming it at this stage. These arguments are made mainly because there is not enough awareness about the critical importance of these s urveys in understanding – in measuring and monitoring – the major concerns of the national economy. The growing literature on this subject (including this paper), which has underlined the importance of TUS data in this context, is bound to defeat this argument.

Finally, it will not be out of place to mention that the mission of the Indian Statistical System, as per the report of the National Statistical Commission (2001), is

To provide reliable, timely and credible social and economic statistics to assist decision-making within and outside the government, and to stimulate research and promote informed debate relating to condition of people’s life.

This mission has led the National Statistical Organisation to c ollect data on different subjects and sectors, the purview of which is determined by the underlying statistical paradigm or “the way the national statistical organisation sees the socio-economic life of people”. This paradigm seems to be influenced by three sets of b iases, namely, the market bias or the “economic” bias (determined by the boundaries of national income a ccounts), the sociocultural bias (including the male biases) and the measurability bias (inclusion of quantifiable or measurable subjects by conventional surveys). As a result, non-SNA activities are excluded from the purview of the paradigm. Looking to the mission of the statistical system, however, this exclusion is not valid. In other words, there is a need to expand the boundary of the national statistical paradigm to include non-SNA activities and personal services.

december 5, 2009

production boundary. They include all delegable production of services not covered under the national income accounts.

7 Personal services are not delegable services, i e, the services that cannot be delegated to others. For example, sleeping, watching the TV, etc.

8 This is observed in South Africa, Benin, Malawi, Morocco, Madagascar, etc in Africa; Thailand, Mongolia, Lao PDR, Bangladesh, Nepal, etc, in Asia; and Argentina, Mexico, Nicaragua in Latin America.

9 Satellite Accounts generally stress the need to expand the analytical capacity of national accounting for selected areas of social concern in a flexible manner, without overburdening or disturbing the central system (UN 2000).

10 The input method uses a replacement wage rate (the wage paid to a person who produces a similar service in the market), which could be a generalist wage rate, the wage rate paid to a housekeeper or a specialist wage rate, or the opportunity cost, i e, the wage rate foregone by the person who is performing the unpaid work. It is observed that the use of a generalist wage rate gives a low value, of specialist wage rates gives a high value, while the opportunity cost gives an unstable value of the work.

vol xliv no 49

11 The technical advisory committee was set up to design the survey, determine the classification and suggest analysis of data. It was headed by I ndira Hirway, the other members included, A C Kulshreshtha, R N Pandey, Anuradha R ajivan and S S Shuka.

12 As regards data collection, India had two alternative methods – the observation method (under which an investigator is expected to observe the r espondents) and the interview method. The first method is not very useful because (1) it tends to make the respondent under observation conscious leading him to behave artificially, (2) if respondents move out of their home, the observer also has to move out, with the result that more than one

o bserver is needed for each household, and (3) the observer cannot be expected to observe respondents during the night and early mornings. It was therefore decided to use the interview m ethod.

13 The United Nations Statistical Division (New York) had organised expert group meetings (EGM) in 1997 and in 2001 to design time use activity classification that can be used in developed as well as developing countries. The present a uthor was a member of these EGMs. ILO had presented a TUS Classification in the 1997 meeting, which was not fully accepted. The 2000 classification, with some modifications in 2003 is still not fully acceptable. There is a need to develop a g lobal classification of time use activities.

14 The objectives of the TAC were (a) to develop a conceptual framework and a suitable methodology for designing and conducting time use studies in India on a regular basis, (b) to infer policy/ p rogramme implications from the analysis of the data, (c) to analyse the time use pattern of the i ndividuals to understand the nature of their work so as to draw inference for employment and w elfare programmes for them, (d) to collect and analyse the time use pattern of people based on the information about their time use, (e) to use the data in generating more reliable estimates on workforce and national income as per 1993-SNA, and in computing the value of unpaid work through separate satellite account.

References

Antonopoulos, Rania and Marzia Fontana (2008): “Gender Awareness and Time Use Data: Why It Matters for Macroeconomic Policies” in Towards Mainstreaming Time Use Surveys in National Statistical System in India, Ministry of Women and Child Development (New Delhi: UNDP and World Bank).

Becker, Gary (1965): “A Theory of the Allocation of Time”, Economic Journal, September.

Bittman, Michael (2008): “Methodological Issues in Conducting Time Use Survey at the Global Level and Emerging Lessons” in Towards Mainstreaming Time Use Surveys in National Statistical System in India, Ministry of Women and Child Development (New Delhi: UNDP and World Bank).

Blackden, Mark C and Quentin Wondon (ed) (2006): Gender Time Use and Poverty in Sub-Saharan A frica, Work Bank Working Paper 73 (Washington DC: World Bank).

Bose, P S (2003): “Employment and Unemployment Surveys of NSS: Estimation of Workforce in Applications of Time Use Statistics”, Central Statistical Organisation, Ministry of Statistics and Programme Implementation (New Delhi: Government of India).

Budlender, Debbie (2008): “Time Use Surveys in Africa”, Feminist Economics, Vol 14, No 3, July, Routlege.

Chakraborty, Lekha (2008): “Statistical Invisibility of Care Economy in India: Evidence from Time Use Data” in Towards Mainstreaming Time Use Surveys in National Statistical System in India, Ministry of Women and Child Development (New Delhi: UNDP and World Bank).

Charmes, Jacques (2004): “Data Collection on the I nformal Sector: A Review of Concepts and M ethods Use Since the Adoption of an International Destinition Towards a Better Comparability of Available Statistics” in Delhi Group. On Informal Sector Statistics, February (Bangkok: ILO).

Jain, Devaki and Malini Chand (1977): “Time Use Survey in Selected Villages in Selected Indian States” (New Delhi: ISST).

Jain, Devaki (2008): “Integrating Unpaid Work into Macroeconomics: Some Indian Experiences” in Towards Mainstreaming Time Use Surveys in N ational Statistical System in India, Ministry of Women and Child Development (New Delhi: UNDP and World Bank).

Kes, Aslihan and Hema Swaminathan (2006) “Gender and Time Poverty in Sub-Saharan Africa” in Gender Time Use and Poverty in Sub-Saharan Africa, World Bank Working Paper 73 (Washington DC: World Bank).

Maitra, Pushkar and Ranjan Rai (2006): “Is there Gender Bias in the Household Time Allocation in a Developing Country? The Indian Experience”, E-Journal of IATUR, Sydney, Australia.

Monsod, Solita Collas (2008): “Integrating Unpaid Work into Macroeconomics: A Short History and the Philippines Experience” in Towards Mainstreaming Time Use Surveys in National Statistical System in India, Ministry of Women and Child Development (New Delhi: UNDP and World Bank).

Motiram, Sripad and Lars Obserg (2008): “Demand or Supply for Schooling in Rural India” in Towards Mainstreaming Time Use Surveys in National Statistical System in India, Ministry of Women and Child D evelopment (New Delhi: UNDP and World Bank).

Nath, S K (2003): “Valuation of Unpaid Household Work and Community Services Using Time Use Statistics” in Applications of Time Use Statistics, Central Statistical Organisation, Ministry of S tatistics and Programme Implementatio (New D elhi: Government of India).

National, Statistical Commission (2001) “Report of the National Statistical Commission Vol I” (New Delhi: Government of India).

Pandey, R N (2000): “Operational Issues in Conducting the Pilot Time Use Survey in India”, Central Statistical Organisation, Ministry of Statistics and Programme Implementation, New Delhi (downloaded from UNSD web site).

– (2008): “Lessons from the Pilot Time Use Survey for Mainstreaming Such Surveys in India” in

T owards Mainstreaming Time Use Surveys in N ational Statistical System in India, Ministry of Women and Child Development (New Delhi: UNDP and World Bank).

Rastogi, Preet (2008) “Rural Child Work, Labour and Daily Practices: Time Use Survey Based Analysis” in Towards Mainstreaming Time Use Surveys in National Statistical System in India, Ministry of Women and Child Development (New Delhi: UNDP and World Bank).

Saha, Vaskar (2003): “Estimation of Workforce Using Time Use Statistics” in Application of Time Use Statistics, CSO, Ministry of Statistics and Programme Implementation (New Delhi: Government of India).

Sharma, Savita and Naresh Kumar (2008): “Valuation of Unpaid Work” presentation made at the international seminar on Towards Mainstreaming Time Use Surveys in National Statistical System in India, Ministry of Women and Child Development (New Delhi: UNDP and World Bank).

UN (1978): “Progress Report on Development of Statistics of Time Use: Report of the Secretary General Submitted to the Statistical Commission at Its 20th Session” (New York: United Nations).

UNDP (2005): “Human Development Report 1995”, HDRO (New York: UNDP).

Economic & Political Weekly

EPW