ISSN (Print) - 0012-9976 | ISSN (Online) - 2349-8846

A+| A| A-

Calculation of Gross Value Added in the Unorganised Sector

The unorganised segment of the economy has been growing steadily in terms of its contribution to the net domestic product, particularly the part of it that falls in the service sector. Measuring the labour input and gross value added per unit of labour input in the unorganised segment are challenging because of the need to take account of workers with multiple economic activities. This paper suggests a way to more accurately compile these statistics on the basis of the enterprise surveys of the National Sample Survey Organisation.

Calculation of Gross Value Added in the Unorganised Sector

An Alternative Approach to Measuring Labour Inputs

The unorganised segment of the economy has been growing steadily in terms of its contribution to the net domestic product, particularly the part of it that falls in the service sector. Measuring the labour input and gross value added per unit of labour input in the unorganised segment are challenging because of the need to take account of workers with multiple economic activities. This paper suggests a way to more accurately compile these statistics on the basis of the enterprise surveys of the National Sample Survey Organisation.

B K GIRI, ASIS ROY, SALIL KUMAR MUKHOPADHYAY

I Introduction

T
he importance of the unorganised segment of the economy is growing over the years. According to the current series (1993-94) of national accounts statistics (NAS), the compilation of one-third1 of the estimate of domestic product pertains mostly to the unorganised segment of nonagricultural production. More than one-third of the domestic product from manufacturing originates from activities carried out in the unorganised segment, which contributes over 5 per cent to the total net domestic product (NDP). More importantly, the production of services in the unorganised segment of the economy constitutes over one-fifth of the total NDP. But there is a scarcity of direct current data for estimating the GDP for this unorganised segment.

The methodology used for compilation of NAS and the sources of data are periodically reviewed in detail to effect improvement in their coverage and reliability on the one hand and to adopt recent developments in the theory and practice of national accounting on the other. Over the years, efforts at methodological improvement in the compilation of NAS, collection of primary data and use of a more appropriate alternative database have enhanced the quality of the NAS considerably. Yet, constrained by the absence of regular current data, particularly in respect of the unorganised segment of the economy, national accounts statisticians are even depending on benchmark estimates and indicators for compilation of national accounts.

II Methods Used in NAS

In India, the labour input method (LIM) is used for obtaining the aggregate estimate of value added for an economic activity or a group of economic activities and is arrived at by multiplying an estimate of labour input going into the process of production and an estimate of value added per unit of labour input for the particular economic activity or group of economic activities. Thus, the essential information required for NAS compilation by the LIM consists of the estimates of labour input and value added per unit of labour input for the unorganised segment.

For the conventional series, the LIM was felt not necessary for NAS compilation up to the 1970-71 series, after which it was adopted using the workforce data of population census (PC) till the current (1993-94) series. In the current series, MWOW2 rates have been used for deriving labour inputs. However, no information on secondary work of the main workers has been collected in PC 2001 and therefore, it is not possible to get MWOW ratios for deriving labour input. The working group set up for revision of the base year for NAS compilation from 1993-94 to 1999-2000 has recommended the count of jobs as a superior measure of labour input and the use of NSS estimates of count of jobs obtained on the basis of usual status3 approach in deriving the labour input and thus, in turn, the labour input ratio – defined as labour input per thousand population. The count of all jobs has been derived by using the estimates of principal and subsidiary capacity employment obtained from the employment and unemployment survey (EUS) of the NSS 55th round (1999-2000). The labour input has been obtained as the sum of estimated number of workers in principal capacity and that of workers in subsidiary capacity separately for each activity category relevant for compilation of NAS. The main advantage of using this method of deriving a measure of labour input from the EUS data is that the distribution of the additional work of the individual workers by activity categories is also generated in the process.

III An Alternative Approach

The system of national accounts (SNA) 1993 also recommends the count of jobs as a superior measure of labour input. The argument is that a part of the working population carries out multiple economic activities. The 1993 SNA further argues that output per job would be an excessively crude measure of productivity and total hours worked would be a better measure of labour input. It also considers the inferior alternative of measuring the labour input in terms of full time equivalent work years.

In view of the methods recommended for obtaining labour input in the NAS compilation for the 1999-2000 series and the recommendations of the 1993 SNA for it, we feel uncomfortable

Economic and Political Weekly September 15, 2007

about using the estimates of workforce obtained on the basis of usual status for deriving the labour input because of the differences in concepts and procedures adopted in identifying workers according to the usual status approach in EUSs from that in enterprise surveys (ESs) of the NSS. Keeping parity with the methodology adopted for estimating value added per unit of labour input for an economic activity or a group of economic activities in the ESs in NSS, we argue in favour of using the daily activity data, collected for each day of the last seven days prior to the date of survey for each individual of the sampled households in the EUSs of NSS, for estimating the count of jobs. The estimate of count of jobs so obtained is expected to be nearer to the estimates one can obtain following the recommendation of the 1993 SNA in respect of “total hours worked” for measuring labour input and also to be in parity with the methodology adopted for estimating value added per unit of labour input for an economic activity or a group of economic activities in the ESs in the NSS. Thus, it is pertinent to examine how the estimates of labour input and value added per unit of labour input for an economic activity or a group of economic activities are obtained in the NAS.

IV Labour Input and Value Added Per Unit of Labour Input

The estimates of labour input may be obtained from PCs and/ or EUSs of the NSSO and those of output or value added per unit of labour input may be obtained from ESs of the NSSO. What is important to note is that an estimate of labour input is also generated in the ESs, inter alia for the purpose of working out an estimate of value added per unit of labour input. As opposed to the estimates of labour input obtained from the PCs and EUSs, which are essentially based on household data, those from the ESs are based on data collected from enterprises. The data from households such as those collected in the PCs and EUSs of the NSSO are usually in terms of employment of persons. It is important to note that the data from enterprises collected in the ESs are usually in terms of jobs. In fact, “worker” in ESs refers to a position or a “job” and a person can pursue one job alongside another. Moreover, in relation to the measurement of productivity of labour, count of jobs is relevant for measuring labour input as well as value added per unit of labour input, or, in the context of ESs, in principle, value added per worker (VAPW). The ES estimates of labour input used for deriving VAPW engaged for an activity, in fact, represent the average number of workers engaged in carrying out regular work in that activity on a typical day of the reference year of the survey, counting both full-time and part-time workers with equal weights.

It may be noted that in the ESs, enterprises are categorised into three types depending upon the number of months of operation during the reference year. Perennial enterprises generally carry out their activities throughout the year, casual enterprises operate their business for some months of the year (not following a regular annual pattern), and others, known as seasonal enterprises, operate during separate seasons only. Workers engaged for a few months of a year in seasonally operated enterprises get the same weight as those in the other enterprises. According to the ES during 2000-01,4 about 92.7 per cent of the unorganised manufacturing units were perennial, 6.2 per cent were seasonal and 1.1 per cent were casual. The distribution of unorganised units among the three types in the service sector (excluding trading units), obtained from the following ES during 2001-02,5 was 98.3 per cent, 1.3 per cent and

0.4 per cent, respectively. The reference period being one year, the typical information collected on average number of workers engaged in an enterprise in the ESs is, theoretically, expected to take into account the seasonality or fluctuations in the job positions in the enterprise during the reference period. This is our main contention against the use of workforce estimates according to the usual status approach. And this objection is stronger for seasonal enterprises. The workforce estimate according to the usual status approach is unable to take care of all kinds of seasonality and labour market fluctuations as it is measured on the basis of the situation obtaining for a person based on a reference period of one year. It is important to mention here that some studies6 have indicated the existence of seasonality in the labour market data. Thus, whatever be the method used for estimating the labour input, it should be free from seasonality and market fluctuations. On this count, we argue in favour of using current activity data (considering both current weekly activity and current daily activity data simultaneously) collected in EUSs to estimate the count of jobs or, in other words, the labour input. Another argument against the estimates based on the usual status approach is that the reference period being the last one year till the date of survey in obtaining the usual status of a person, it describes the job market beginning with 365 days before the survey period and therefore, loses its relevance, to some extent, in measuring labour input corresponding to the survey period in a growing economy. On the other hand, the current activity data refer to the situation obtained on the basis of each day of the last seven days prior to the date of survey, and correspond more or less to the survey period.

V Daily Activity Data in EUS and SNA Requirement

For each individual, irrespective of age and sex, the various activities performed or inactivity in each of the last seven days prior to the date of survey are recorded in quantitative terms of “half” or “full” intensity. Thus, while on a particular day, a person may have any number of activities, only the particulars relating to one or at most two activities identified on the basis of priority-cum-major-time criterion are recorded. Generally, a current activity which is pursued for more than

Table 1: Labour Input Ratio according to Current Status Approach Separately for Each Sex and Sector: NSS 55th Round (1999-2000)

Category of Persons Labour Input Ratio wpr (According to cws)
(1) (2) (3)
Rural male 650 510
Rural female 294 253
Rural person 476 384
Urban male 645 509
Urban female 155 128
Urban person 410 327
Rural + urban male 649 510
Rural + urban female 259 222
Rural + urban person 459 370

Economic and Political Weekly September 15, 2007 one hour but less than four hours is considered to have been pursued with “half” intensity. If it is pursued for more than four hours, the activity is considered to have been pursued with “full” intensity. On the other hand, if four hours or more are spent on each of two or more activities, then at the most two activities are chosen on the basis of priority (employed over unemployed and unemployed over out of labour force)cum-major-time criterion and recorded. It is important to note that the various activities of a person are distinguished one from another by the status-cum-industry (at two digit level of NIC) of the activities performed in a day. The different work activities in terms of status-cum-industry (at two digit level of NIC) performed during the seven days of the week by a worker according to the current weekly status have been considered as the jobs held by such workers according to the current status approach. This information in respect of status-cum-industry has been used for counting the jobs classified according to sex, industry and rural/urban sector. The job count from this data is analogous to hours of work (as recommended in SNA 1993) and representative of average number of workers engaged in an enterprise in the ESs.

VI Estimate of Labour Input Ratio from Current Activity Data

The labour input ratios obtained by using the current activity data are given separately for each sex and sector at the national level in Table 1. Data used for this exercise relate to the EUS survey of the NSSO during 1999-2000. Out of every 1,000 persons in the country, about 370 were found to be workers according to current weekly status (cws)7 and they worked, on an average, against 459 job positions – termed as labour input ratio. In other words, each worker was engaged in 1.24 jobs – call it job worker ratio. Either of the ratios, labour input ratio or job worker ratio, can be used, the first on the projected population and second on the workforce, to get the labour input required for NAS compilation. Similar statistics in respect of labour input ratio are given in Table 2 separately for each tabulation category of activities (NIC 1998) side by side with the ratio based on the usual status approach for NAS compilation (1999-2000). The two sets of estimates, at the broad level, reveal coherent magnitudes and directions in respect of the approaches and methodologies. The NAS estimates of worker population ratio (wpr: number of workers per 1000 population), labour input ratio and job worker ratio are found to be 397, 485 and 1.22, respectively.

It is important to note that the labour input ratio according to the current status approach is lower than that according to the usual status approach in the case of the primary sector but higher in the case of the secondary and tertiary sectors (Table 3). The labour input ratio according to the current status approach is expected to behave in a similar way for the unorganised activities in the secondary and tertiary sectors, as it is obtained by the residual method by subtracting the labour input for the organised sector from the total labour input.

The overall reduction in labour input ratio according to the current status approach compared to the usual status approach may be due to the fact that the labour input ratio obtained from the current status approach takes into account the seasonality and fluctuations in the labour market. The increase in the labour input ratio in the secondary and tertiary sectors along with decrease in the primary sector according to the current status compared to the usual status approach may be due to the fact that some persons assigned a primary industry (for a period of 365 days) according to the usual status approach may not have worked in the primary sector throughout the year. Some of them may have worked in the secondary/tertiary sector during the off season in the primary sector or when they did not have adequate employment in the primary sector and this is reflected in the labour input ratio obtained in the current status approach.

VII Limitations in Current Activity Status Data

It may be mentioned that there are some limitations in estimating the count of jobs from the data available according to the current activity status approach due to the procedures8 adopted for recording the status-cum-industry of each individual

Table 2: Labour Input Ratio by Tabulation Category1 as Obtained from Usual Status and Current Status Approach: NSS 55th Round (1999-2000)

Tabulation Category Labour Input Ratio Usual Status2 Current Status

(1) (2) (3)

A+B 320 279 C 2 2 D 46 51 E 1 1 F 20 19 G 38 43 H 5 5 I 15 16 J 2 3 K 3 3 L 10 12 M 8 10 N 3 3 O 11 10 P 2 2 Q 0 0 All India 485 459

Notes: 1 Description of tabulation categories: A, agriculture, hunting and forestry; B, fishing; C, mining and quarrying; D, manufacturing; E, electricity, gas and water supply; F, construction; G, wholesale and retail trades; repair of motor vehicles, motor cycles and personal and household goods; H, hotels and restaurants; I, transport, storage and communications; J, financial intermediation; K, real estate, renting and business activities; L, public administration and defence; compulsory social security; M, education; N, health and social work; O, other community, social and personal service activities; P, private households with employed persons; Q, extra-territorial organisations and bodies.

Source: 2 Report of the Working Group on ‘Workforce Estimation for Compilation of National Accounts Statistics with Base Year 1999-2000’.

Table 3: Labour Input Ratio by Broad Categories of Industry as Obtained from Usual Status and Current Status Approaches: NSS 55th Round (1999-2000)

Broad Categories of Industry (Tabulation Category) (1) Labour Input Ratio Usual Current Status Status (2) (3) Per Cent Change in Current Status over Usual Status (4)
Primary (A-C) Secondary ( D-F) Tertiary (E-O) All (A-Q) 322 67 94 458 281 71 105 459 -12.73 5.97 11.70 -5.36

Economic and Political Weekly September 15, 2007

in EUSs of the NSSO. In the current daily activity status approach, “full” intensity of work is recorded corresponding to the status-cum-industry for regular wage/salaried employees, whatever other jobs they perform on a particular day, and thus, their secondary work, if any, makes no contribution to the job counts. The secondary work of the regular wage/salaried employees has been estimated through an indirect method, using the proportion of workers among the regular wage/ salaried employees according to usual principal status having subsidiary jobs (see Table 4). This method of adjustment is found to be quite satisfactory from Table 5. From Table 5, it is observed that almost all workers categorised as regular salaried employees (about 98 per cent) in a reference week are also found to be regular salaried employees in the usual principal status. Thus, the adjustment for the secondary jobs of the regular salaried employees in a reference week based on the information on usual principal status workers having subsidiary jobs is quite acceptable.

Even after this adjustment, there are still some reasons to believe that the number of jobs counted according to the current activity status approach remains an undercount. This is also due to the procedures in EUSs adopted for recording the intensity of work for a person who performed two or more jobs in a day – one for a duration of four hours or more, and others for one to less than four hours. In this case, the procedure in the EUSs is to record ‘full’ intensity against the first job, ignoring the remaining.

Third, for NAS purposes, it is essential that the labour inputs are estimated for the compilation categories of the activities. This is not directly possible from the daily activity status data. The data in respect of daily activity status are collected at the two digit level of the NIC and hence, the distribution of jobs at the compilation categories of the activities cannot be directly obtained. On the other hand, based on the daily activity status-cum-industry, a five digit industry of activity is ultimately assigned to the working persons in the current weekly status based on priority-cum-major-time criteria. As a crude measure, the labour input ratio at the two digit level of industry obtained from daily activity data, as explained above, may be applied to the distribution of workers (according to current weekly status approach) for compilation categories – assuming that the labour input ratio at the two digit level of industry holds good for all the groups, classes and sub-classes under that industry division.

VIII Some Suggestions

Here, we want to address a few small problems that are encountered in estimating the labour inputs using current activity status data. First, the data in respect of daily activity status may be collected at the five digit level of NIC also, as is done in the usual or current weekly statuses, but change in the status-cum-industry may be restricted to status-cum-twodigit-level of NIC as is done now in EUSs. This is only to facilitate the estimation of labour inputs at the compilation categories of the activities required for NAS compilation. In fact, this is now done, in general, in deciding the current weekly status and industry of activity for those who perform a single activity throughout the week. In the case of those who perform multiple jobs in the week, five digit industry of activity is recorded against one of the work activities based on priority-cum-major-time criteria. The value 1.24 of the job worker ratio implies that, in about 19 per cent cases, the five digit industry code is to be recorded additionally, to comply with this suggestion.

Second, although there are problems in estimating the multiple jobs performed by the regular wage/salaried persons in the reference week, we still prefer to retain the same procedure of recording “full” intensity for each day of the reference week for the regular wage/salaried persons. The reason is that the total number of mandays worked by them in the reference week is needed for wage calculation. We prefer to borrow the adjustment factor using the usual principal status cross-classified by usual subsidiary status for the regular wage/salaried employees. The adjustment factor obtained, as explained earlier, we feel, is perceptible and fairly good.

Third, we suggest that at least two jobs in a day be recorded in terms of status-cum-industry, if two or more jobs are performed – one for four hours or more and the others for one to four hours, with “half” intensity each by using priority-cummajor time criteria. This procedure will enhance the scope of recording multiple jobs, at the most two, performed in a day and thereby, estimating the labour inputs more precisely. Note that the procedure will not disturb the estimate of mandays worked in the economy; rather, it will result in meaningful estimates of wage rate (productivity of labour).

IX Conclusion

It is evident that the labour input ratios obtained from the current activity status data of the EUS of the NSS are expected to provide a superior measure of count of jobs as defined in the enterprise surveys of the NSS, and this is also in close agreement with the SNA 1993 in respect of hours of work.

Table 4: Number of Jobs Per Regular Wage/Salaried Employee according to Usual Principal Status: NSS 55th Round (1999-2000)

Category of Persons Jobs Per Regular Wage Salaried Employee

(1) (2)

Rural male 1.1886 Rural female 1.1059 Urban male 1.0228 Urban female 1.0238

Table 5: Percentage of Regular Wage/Salaried Employees (Activity Status: 31) according to Usual Principal Status Who Had Regular Wage/Salaried Employment in Current Weekly Status

Category of Persons Percentage

(1) (2)

Rural male 96.99 Rural female 96.27 Rural person 96.88 Urban male 98.72 Urban female 98.33 Urban person 98.66 Rural + urban male 98.05 Rural + urban female 97.55 Rural + urban person 97.97

Economic and Political Weekly September 15, 2007 We feel that the collection of current daily activity data at five digit level of industry is quite possible as the increase in the workload in conducting the survey is marginal. If the EUSs are conducted annually for a larger sample of households, one can use the current activity status data obtained from such surveys to update the labour input ratio regularly and this will do away with the stage of extrapolating the labour input ratio during the period between two quinquennial EUSs for NAS compilation.

EPW

Email: bimalgiri1953@yahoo.co.uk

Notes

[This paper expresses the authors’ personal views and does not represent the opinions of the organisation with which they are associated.]

1 Report of the Working Group on ‘Workforce Estimation for Compilation of National Accounts Statistics with Base Year 1999-2000’.

2 Apart from the count of “main” and “marginal” workers, the PC 1991 also provided count of MWOW, i e, “main” workers who were engaged in an economic activity other than the “main” work. The MWOW ratio was defined as the ratio of MWOW to total workforce (“main” workers plus “marginal” workers). The labour input was then derived by inflating the estimated workforce with the MWOW ratios.

3 The estimate of workforce according to the “usual status” approach represents the number of persons engaged in some economic activity, in either principal or subsidiary capacity, or both, during the 365 days preceding the date of survey. A person engaged in some economic activity “more or less regularly” for a relatively shorter period during the reference period of the last 365 days is treated as having subsidiary capacity work. A person may be engaged for a relatively longer period during the last 365 days in economic/non-economic activity and for a relatively shorter period in “another economic activity”, and a person may be pursuing one economic/non-economic activity almost throughout the year in the principal usual activity status and also pursuing alongside “another economic activity” in a subsidiary capacity – be it almost throughout the year or for a relatively shorter period. Here, “another economic activity” means either the change of activity status or change of broad industry of activity (at the two digit level of the NIC).

4 NSS Report No 478: Unorganised Manufacturing Sector in India, 20002001.

5 NSS Report No 483: Characteristics of Enterprises: Unorganised Service Sector in India, 2001-2002.

6 Report of the Group on Generating Quarterly Labour Market Data under SDDS, Ministry of Statistics and Programme Implementation, April 2005.

7 According to the cws approach, a person is considered a worker if he/ she has performed any economic activity for at least one hour on any day of the reference week; his/her current weekly status is obtained from the daily activities performed or in activity on each day of the last seven days prior to the date of survey.

8 Refer to Instructions to Field Staff, Vol I, NSS 55th round.

Economic and Political Weekly September 15, 2007

To read the full text Login

Get instant access

New 3 Month Subscription
to Digital Archives at

₹826for India

$50for overseas users

Comments

(-) Hide

EPW looks forward to your comments. Please note that comments are moderated as per our comments policy. They may take some time to appear. A comment, if suitable, may be selected for publication in the Letters pages of EPW.

Back to Top