Introduction

 

At the end of 2003 the Federal Bureau of Statistics of Pakistan (FBS, herewith) has completed the migration of their National Accounts to the new base year 1999-2000 and has upgraded its compilation methodologies, following, where possible, the recommendation of 1993 SNA. Within the re-basing exercise the FBS has requested UNESCAP to assist them in developing appropriate techniques to estimate the contribution of informal sector to GDP, using the existing data sources.

 

In Pakistan a direct estimation of the value added of the informal sector is not feasible at the moment, given that no special enterprise (or mixed) survey has been carried out so far. Indirect estimation approaches, however, can be tested given that information on employment and hours worked in the informal sector are collected in Pakistan since 1996-97 through the Labour Force Survey (LFS). The “labour input method”, in particular, has been studied and applied at international level[1] in order to derive proxy estimates of this component of the Non-observed economy. This methodology is based essentially on three steps: obtain estimates of labour input in the informal sector by economic activity from household surveys (generally the LFS); derive estimates of output and value added per unit of labour input with the same economic activity breakdown for enterprise belonging to the informal sector; multiply the labour input estimates by the per unit ratios to get output and value added of the informal sector by industry. The implementation of this approach, therefore, requires data support also from establishments’ surveys in order to obtain the essential information on labour productivity.

 

This paper describes the main statistics on informal sector collected in Pakistan and the major issues involved in integrating this component of the economy in the compilation of GDP. In section 2 the estimates of employment in the informal sector based on the LFS are presented, along with some critical consideration on the accuracy of LFS results. Section 3 contains a brief description of the methodology followed for the estimation of Value Added of informal sector and an assessment of the feasibility of its adoption for the compilation of NA in Pakistan. Some concluding remarks follow.

 

1.                Employment in the informal sector

 

The starting point in the exercise aimed at integrating informal sector activities in the compilation of GDP is the presence in Pakistani statistical data collection system of a well established module on informal sector attached to the questionnaire of the LFS. Information on the characteristics of the enterprise concerning legal status, accountancy and employment size is collected since 1996-97, when a new module on informal sector activities was incorporated in the questionnaire of the LFS.

 

Estimates of employment in the informal sector by industry (2 digit) and status in employment can be obtained from the LFS for the base year (1999-2000) and for a few contiguous years (1996-97, 1997-98, 2001-2002). Data for the missing years can be estimated at the industry level either with a linear interpolation or identifying, if possible, the trend of the series of informal employment.

 

In general, the labour input approach foresee that employment estimates should be complemented by information on hours actually worked in order to convert stocks into flows and obtain estimates of the actual amount of labour input utilized in the production process over a determined period of time. However, as we will see in the next section, the lack of information concerning hourly productivity makes this exercise useless.

 

Before providing a detailed description of the characteristics of people engaged in the informal sector and the evolution of this aggregate over the last five years according to the LFS, the national definition and the questionnaire of the LFS are discussed, highlighting their main limitations in terms of international comparability; moreover, the accuracy of LFS estimates is analyzed, checking the internal consistency of the information collected and comparing the results with external data sources.

 

1.1.     National definition of informal sector

 

The operational definition of informal sector adopted in the LFS is based on two criteria: the enterprise has to belong to the household sector (in other words, only private unincorporated enterprises are included in the informal sector) and its size has to be below a certain threshold, established in terms of number of persons engaged. The first criterion is essential for the definition, whereas the second is applied only to enterprises that employ permanent employees. The criterion of registration of the enterprise as a corporation or with the tax authorities, instead, is not used “given the ambiguity of the registration criterion in the circumstances of Pakistan, and the difficulty of obtaining reliable response on this item from household members” Following the international guidelines, all household enterprises engaged in agricultural activities are excluded from the definition. Paid domestic employees and family members producing goods for own final uses, instead, are not excluded in principle, because of the old classification of economic activities (ISIC Rev. 2) used at very low level of detail (2 digit) that prevents these categories to be extracted from the aggregate. The other main deviation from the recommendations of the ICLS is that informal activities carried out as a secondary job are not included in the aggregate as questions on the informal sector refer only to the main activity of the persons employed.

 

The module on informal sector activities included in the LFS questionnaire comprises 7 questions:

 

q10:           What kind of enterprise?

q11:           Does the enterprise keep written accounts?

q12:           How many persons are engaged in the enterprise?

q13:           Are there any regular paid employees in the enterprise?

q14.1:       If yes, number of males?

q14.2:       f yes, number of females?

q15:           Where did … carry out the work?

 

Questions 10 and 11 are intended to single out all the persons employed in private unincorporated enterprises: to begin with, all the persons working for the Government (at Federal, Provincial or Local level) or for public and state-owned enterprises, corporations and cooperatives are excluded; then, only the remaining enterprises that do not keep written accounts can be classified as private unincorporated enterprises because a financial separation of the production activities of the enterprise from the other activities of its owner(s) is not possible.

 

Question 13 is meant to distinguish between own-account enterprises and small enterprises on the basis of the presence of regular paid employees: as already anticipated the employment size criterion applies only to the second group of enterprises. Therefore all household enterprises owned and operated by own-account workers are included in the informal sector irrespective of the size of the enterprise (informal own-account enterprises); small enterprises are included only if the total number of persons engaged in the enterprise (as a whole) is below 10 units (question 12).

 

The definition of the informal sector adopted by Pakistan is not completely coherent with the harmonised definition resulting from the recommendations made by the Delhi Group: in particular, the criterion of registration is not used; the size of the enterprise is established on the total number of persons engaged; the size limit is set at 10 people; households employing paid domestic employees are included in the definition.

 

In addition to that, the present questionnaire of the Pakistani LFS doesn’t allow a full implementation of the harmonised definition, mainly because the question on registration is not present and because the classification of economic activities is not up to date. Information on the exact number of permanent employees working for the enterprise have started to be collected with the 2001-2002 edition (with the introduction of questions 14.1 and 14.2), allowing to apply the size limit criterion and to derive estimates closer to international standards. The reduction in the cut-off size at 5 employees, however, does not imply a major change in the size the informal sector as the total number of persons engaged decreases roughly by 1% only.

 

The other question introduced in 2001-2002, concerning the place of work (question 15), is not relevant for the definition of the informal sector but can be used to identify certain subgroups of workers (home-based workers, street vendors) within the main aggregate.

 

1.2.           Quality issues

 

In Pakistan, as we will see in the next paragraph, the share of employment of the informal sector is quite high overall and covers nearly the totality of the persons engaged in industries like Construction, Wholesale and Retail Trade and Hotels and Restaurants. Even accounting for differences of the national definition of informal sector from international standards, the share of employment in Pakistan is higher than most of the other Asian countries for which data are available. This raises some concerns about the accuracy of the LFS results.

 

One of the possible sources of inaccuracy of the LFS estimates is that also paid employees, who may not have exact information on the characteristics of the enterprise, are asked questions about the legal status, the accountancy practices and the employment size of the firms where they work. This problem may be exacerbated in case of proxy respondents, i.e. when some other member of the household is participating in the interview on behalf of the relevant person. It is unclear, though, if this kind of measurement error is systematic and which is the direction of the bias.

 

In general there are two main ways of detecting measurement errors: the first one is through the comparison with external data sources, performed at aggregate or at unit level; the second one is through checks of internal consistency of the information collected.

 

The only external source, whose employment figures can directly be compared with the LFS, is the Survey on Small Scale and Household Manufacturing Industries in Pakistan. This survey, carried out approximately every 3/4 years and last time in 1999-2000, covers all the enterprise with less than 10 employees belonging to Manufacturing: limiting the analysis to this industry, therefore, the SSHMI is characterized by a larger universe than the LFS because all the enterprises with less than 10 employees are included in the population, regardless of their legal status and accountancy practices; moreover the size limit is fixed with reference to the number of employees and not of total persons engaged. As expected, informal sector employment in the LFS is lower than total employment in the SSHMI survey in 1999-2000, but the discrepancy between the two figures is just 10%.

 

Alternative estimates of employment in the informal sector to validate the LFS results can be derived only for Manufacturing and Mining and Quarrying through the comparison of 2 different data sources, the LFS, on one side, the Censuses of Manufacturing Industry and the Censuses of Mining and Quarrying, on the other side. The CMI and the CMQ are conducted on a yearly basis with a complete coverage of all registered enterprises with at least 1 employee, irrespective of their size. To the extent that registered enterprises constitutes separate legal entities with a complete set of written accounts, there should be no overlap between the informal sector defined according to the LFS criteria and the aggregate of registered enterprises with less than 10 employees. The residual obtained comparing total employment between supply side and demand side surveys constitutes an indirect estimate of employment in the informal sector that can be compared with the direct estimate coming from the LFS. There are a number of reasons why these two estimates may differ. In general, the indirect procedure can be expected to produce higher figures, as it may include people employed in the underground sector (for example, non registered employees working in firms of medium and big size). Moreover, as the number of workers reported in the establishments’ surveys relates to the number of jobs, and it may be possible for a person to have more than one job, the number of persons employed may be less than the total number of jobs/workers reported in establishment surveys. The reliability of the indirect estimates will depend also on the capacity of the LFS in providing accurate distribution of workers by industry, to correspond to the breakdown taken from the establishments’ surveys.

 

Table 1:   Missing information on the characteristics of the enterprise by status in employment

 

Source: Pakistan Labour Force Survey 2001-2002

 

A better understanding of the quality of the answers to the questions on the characteristics of the enterprise can be grasped examining the internal consistency of the values reported in the survey as some the responses to the questionnaire are not cross-checked by the FBS as part of the usual editing process. The main missing information and inconsistencies in relation to the status in employment are the following (see tables 1 and 2):

 

1.      1.8% of self-employed people against 6.7% of employees couldn’t identify one of the 9 typology provided at the question on legal status of the enterprise (question 10);

 

2.      Nearly ¼ of the employees don’t have information on the accountancy practices of their enterprise (question 11), whereas only 4.3% of the self-employed don’t answer to this question. Overall, 14.6% of the employed in the informal sector don’t know if their enterprise keeps written accounts and are included in the aggregate without further probing on the characteristics of the firm. The vast majority of these people (82.3%) are employees.

 

3.      The proportion of respondents declaring their firm has no regular paid employees on its payroll (question 13) is 34% for regular paid employees (but 93.9% for casual employees) and 48% for employers; on the other hand, only 3.4% of own-account workers’ reply affirmatively to this question. Using the information on status in employment instead of the question on the presence of permanent employees in order to distinguish between own-account enterprise and small enterprise, therefore, would lead to a change in the size of the informal sector.

 

4.      Overall 5.5% of the respondents provide a number of employees (question 14.1 and 14.2), bigger than total employment of the same firm (question 12). In this case the quota of inconsistent answers doesn’t vary significantly according the status in employment. The possible influence of these errors (enterprise classified as informal, i.e. with less than 10 persons engaged, but having more than 10 employees) on the size of informal sector is negligible.

 

Table 2 – Inconsistencies in the replies to the questionnaire by status in employment.

 

Source: Pakistan Labour Force Survey 2001-2002

 

Given the lower quality of employees’ replies to the questions on the characteristics of the enterprise, an alternative estimate of the size of informal sector based on data internal to the LFS can be derived taking into account only the information collected on employers and own account workers. Due to the limitations of the questionnaire, though, this alternative estimate can be calculated only for employees working in small enterprises, which represent only the 21.9% of the total employees engaged in the informal sector. According to the replies given only by employers and own account workers, the number of permanent employees of the informal sector is 10.8% lower than the official estimate. On the other hand, the total number of employees of small enterprises estimated with this alternative methodology is much higher than the official figure (and more precisely 39.1% bigger); as a result, the share of employees of the informal sector for small enterprises drops from 59.3 to 38%.

 

It seems therefore that according to employers the size of the firms is significantly higher than what employees perceive, in contrast with the general view that employers tend to underreport the number of employees (above all casual employees). Part of these discrepancies, on the other hand, can be accounted for by the high sampling error related to the very low number of employer that have been surveyed in the LFS.

 

1.3.     Characteristics of workers of the informal sector

 

According to the latest edition of the LFS (2001-2002) the employment of the informal sector in Pakistan is estimated at 14.746 million, representing a share of 66.5% over non-agricultural employment and 38.5% over total employment (Table 4).

 

Table 4 - Pakistan. Composition of employment in the informal sector

 

Total employment	38,303,888
Total non-agricultural employment	22,181,781
Percentage of total employment	57.9
Non-agricultural employment in household enterprises	16,546,537
Non-agricultural employment in household enterprises with no written accounts	14,906,687
Employment in informal sector	14,746,040
Percentage of total employment	38.5
Percentage of non-agricultural employment	66.5
- Employment in informal small enterprises	1,838,181
- Percentage of total informal employment	12.5
- Employment in informal own-account enterprises	12,907,859
- Percentage of total informal employment	87.5

Source: Pakistan Labour Force Survey 2001-2002

 

This aggregate comprises mainly workers engaged in informal own-account enterprises (87.5%), while the other segment, represented by workers of informal small enterprises, accounts for only 12.5% of the total. Consequently, the self-employed constitutes the larger component of informal sector employment (54.4%), whereas the employees are relatively less numerous and are concentrated mainly in informal own-account enterprises, with casual relationship with their employer. In the formal sector, instead, the bulk of the workforce is represented by wage and salary earners (87.9% of total employment in the formal sector).

 

As in most of the other countries, the size of the informal sector is very large in Construction, Wholesale and Retail Trade and Hotels and Restaurants (with a share over total employment higher than 85%), industries where employment follows frequently a seasonal pattern and the enterprises have less incentives to hire workers on a regular basis (Table 5). Overall the weight of these sectors is more than three times higher in the informal sector than in the formal one. On the contrary, the phenomenon is less widespread in Manufacturing (with a share of 56.9%), and in Business and in Other services (28.8 and 47.2%, respectively); these industries, therefore, account for a relatively small proportion of total informal employment, as compared to the one they cover in the formal sector. The differences in the composition by employment status and industry affect many other characteristics of the workforce of the informal sector, as well as its working conditions.

 

Concerning demographic variables, large differences between the internal distribution of formal and informal sector employment can be found by educational level and to a less extent, by age group and district, whereas the proportion of women to men is nearly similar in the two segment of non-agricultural employment (Table 6): this result, however, can be influenced heavily by large undercounting of women participation to the labour market in the LFS. As for the age variable, a U shaped pattern can be detected (Table 5), with a maximum share of informal sector for the young workers and the most aged ones (82.5 and 80.6%, respectively), and a minimum value for the individuals in the 30-39 years bracket; overall, however, the average age of the workers of the informal sector is considerably lower than of the ones of the formal sector. The size of informal sector decreases steadily as the educational attainment increases, with a maximum value for illiterate workers and a minimum for people with secondary or tertiary level of education; overall, nearly 45% of workers of the informal sector have never received any formal education.

 

As for geographical variables, the lowest share of informal sector has to be found in Sindh and Balochistan, while the maximum is in Punjab; in this district, in particular, is concentrated over 2/3^rd of total employment of the informal sector, against only half of the formal one. By location of work, informal sector is relatively more important in rural than in urban areas (72.6% against 62.6%), even if the employment of informal enterprises located in the countryside weight only 42.4% over total employment of the informal sector. Moreover, nearly half of this component works in “informal places”, either in private dwellings or on the street, while the large majority of the employment of formal enterprises works in establishments.

 

 

 

  

 

Source: Pakistan Labour Force Survey 2001-2002

 

 

 

Source: Pakistan Labour Force Survey 2001-2002

 

The LFS questionnaire collects also some information that can be used to analyze the working conditions of the people engaged in informal enterprises as compared to the formal ones. A set of questions ascertain the amount of weekly working time and additional information are collected about the willingness to work additional hours and the presence of job search activity. Moreover, by comparing the current employment status with the one of the previous year it is possible to build an indicator of job mobility. Lastly, but only for the employees, the weekly and hourly wages can be compared the two components of non-agricultural employment.

 

On average, men employed in the informal sector work 50.5 hours per week, 2 hours more than the corresponding population of the formal sector, while women employees of informal enterprises work 38.2 hours, nearly 2 hours less. In particular, it is in Wholesale and Retail Trade and in Manufacturing, where the differences in working times of the 2 components of female employment are considerably wide. Weekly working time of women wage earners is largely inferior to the one of men, both in the formal and the informal sector; in this segment of non-agricultural employment, though the differences by sex are greater.

 

Table 7 -   Pakistan. Average number of hours of work per week in the informal sector, by industry and sex of the employee
Industry	Values		Diff. with Formal Sector			Hours worked by women in the informal sector as a % of men’s
	Men	Women	Men	Women
Manufacturing	50.8	35.8	-0.2	-10.3		70.5
Construction	45.0	46.7	-2.9	. .		103.8
Wholesale & Retail Trade	54.3	42.6	0.8	-24.0		78.5
Hotels & Restaurants	58.4	52.9	2.0	. .		90.6
Transport & Communications	56.2	49.5	3.7	2.8		88.1
Business services	47.7	. .	1.1	. .		. .
Other services	51.3	40.1	6.1	1.2		78.2
Total	50.5	38.2	2.0	-1.9		75.6

Source: Pakistan Labour Force Survey 2001-2002

 

Employees working less than 35 hours a week are concentrated among women, mainly in the informal sector. The relatively lower working time leads women to search for additional hours of work. The quota of underemployed is only slightly higher in the informal sector as compared to the formal one. The difference in the percentage of employed seeking a new job with higher working hours is even lower between formal and informal sector.

 

Large disparities by sex can be found also concerning compensation of employees. On average, women working in the informal sector earn less than half of the weekly salaries of men. Again, the differences between men and women are higher in the informal sector than in the formal one, as the wage premium for being employed in formal enterprises is higher for the female workforce.

 

Table 8 - Pakistan. Average wages per week in the informal sector by                 industry and sex of the employee
					Women’s avg. wages in the informal sector as a %of men’s
Industry	Values		Informal/Formal Ratio
	Men	Women	Men	Women
Manufacturing	706	290	70.2	37.7	41.1
Construction	684	477	90.0	. .	69.7
Wholesale & Retail Trade	576	512	57.1	84.1	88.9
Hotels & Restaurants	720	409	72.4	. .	56.8
Transport & Communications	818	608	68.3	45.7	74.3
Business services	1,124	. .	43.4	. .	. .
Other services	653	337	59.6	37.8	51.6
Total	694	319	62.7	35.8	46.0

Source: Pakistan Labour Force Survey 2001-2002

 

These wage inequalities can be explained only partially by the differences in working time. In fact, hourly wages in the informal sector for female employees are 39.2% higher than for females. A positive wage differential in favor of women can be found only in Wholesale and Retail Trade, an industry that is characterized by a relatively low working time for female employees.

 

Source: Pakistan Labour Force Survey 2001-2002

 

Mobility in the labour market is significantly affected by more casual working arrangements and, more generally, by the inferior working conditions that characterise jobs in the informal sector. In this segment of the economy the percentage of people whose job has lasted more than one year is 86.1%, whereas in the formal sector this indicator reaches 94.4%. Job turnover explains much of the differences. Among people employed in the informal sector mobility from unemployment to employment (and vice-versa) is higher than in the formal sector.

 

1.4.           Employment trend in the informal sector

 

In Pakistan the employment of the informal sector has increased over the five year period 1996-97/2001-2002 at a pace significantly faster than the one of the formal sector (the average annual growth rate being 1.8 and 1.6%, respectively). During the first three years the two components followed a counter-cyclical pattern, leaving total non-agricultural employment approximately unchanged; in the following biennium, instead, both have shown a strong rebound with a slightly higher growth rate for the informal sector. The total number of person engaged in this sector has in fact jumped from 12.318 million in 1996-97 to 14.746 million in 2001-2002. Over the same period, non-agricultural employment has at first slightly decrease (from 19.062 million in 1996-97 to 18.7340 million in 1999-20002) and has recovered significantly afterwards, getting to 22.182 million units. The size of the informal sector has therefore risen from 64.6% in 1996-97 to 66.5% in 2001-2002.

 

Fig. 1 - Employment trends in formal and informal sectors

 

 

The industries that have experienced the most rapid growth of employment in the informal sector over the whole period are manufacturing (2.2% of average annual growth rate) and business services (2.3%); in these industries the size of the informal sector has increased considerably, especially in business services, where the formal component of employment has fallen at the same time. Above the average has also been the expansion of informal sector employment in transport and communication (1.9%), whereas relatively slow has been the rate of increase shown by Construction and Hotels and Restaurants, two among the three industries with the highest share of informal sector employment (just below 90%).

 

Source: Pakistan Labour Force Survey 2001-2002

 

2.                Estimation of Value Added of informal sector

 

2.1.           Labour productivity in the informal sector

 

Following the labour input methodology, the second step consists in deriving appropriate estimates of output and value added per worker from enterprise surveys. In this respect, the situation of the existing statistical data collection system in Pakistan is particularly poor. With reference to the fiscal year 1999-2000 a large survey targeted to collect data on small scale and home based enterprises was carried out only for manufacturing (SSHMI survey). For the rest of the economy, small special surveys were conducted on a number of industries, but information on labour productivity by size of enterprise has never been published. The acquisition of the files of data at unit level for these special surveys is the basic requirement to complete the project.

 

The only other establishment survey with a broader activity coverage than just Manufacturing was conducted by the FBS in 1993 (on fiscal year 1992-93). As for the size the survey was intended to cover all registered and unregistered establishments having less than 50 employees. The results of the ISSMI have been used extensively by the FBS in order to compile the N.A., but they were never disseminated. Even if these data were made available with the required detail by firm size, however, quite strong assumption on productivity growth and inflation should be made to derive the estimates of output and value added per worker for year 1999-2000.

 

Given the lack of data on Pakistani economy, an alternative exercise has been carried out using the corresponding figures for the Indian economy collected through the 55^th round of the NSS. As shown in table 11, even after having converted the original data from Indian Rupee to Pakistani Rupee using the Purchasing Power Parity conversion factor for year 2000, the GVA per worker of the informal sector obtained from the Indian survey in Manufacturing is much lower than the one taken from the SSHMI survey; moreover, the adjusted Indian GVA per worker is even lower in most industries than the annual salary of employees of the informal sector coming from the LFS of Pakistan. Assuming that the industrial structure of the two economies doesn’t differ in the relative capacity of producing income, the productivity differentials between industries obtained for India have been applied to the GVA per worker of Manufacturing surveyed in Pakistan. These estimates of GVA per worker have then been applied to the employment in the informal sector by industry to obtain the total value added produced in the segment of the Pakistani economy.

 

 

2.2.           Contribution of the informal sector to the GDP

 

In table 12 the contribution of informal sector to total GDP for fiscal year 1999-2000 is presented. No value added is produced in the informal sector in industries like Agriculture and Public Administration by definition. Ownership of dwellings’ income is inconsistent with informal sector employment, as the imputed rents cannot be considered as corresponding to work generating labour compensation or mixed income. For the remaining industries, the production of the informal sector has not been added to the revised estimate of GDP for year 2000 because the methodology used in the compilation of the new set of National Accounts has not been released yet and it is not possible to check if this operation would determine a double counting of some economic activities and over-estimation of GDP. For manufacturing, in particular, the output coming from small scale and home based enterprises is already taken into account in the new value added estimates. Following this approach some problems arise only in the construction industry where the estimates of the value added produced in the informal sector are considerably higher than total GDP. It is likely therefore that the revised NA estimates do not cover all the production activities carried out in this industry: it should be noted, incidentally, that construction is one of the few sectors where the re-basing exercise has led to a downward revision of the “old” estimates of value added with base year 1980-81.

 

The contribution of informal sector to GDP is substantial in Wholesale and Retail Trade and Hotels and Restaurants, accounting for over half of the total value added of the industry; quite large is also in Transport & Communications and especially in Social, Community & Personal Services if compared to the relatively low share of employment in the informal sector typical of this industry. For Manufacturing the value added produced in the informal sector represents only 15% of total output, a percentage considerably lower than the corresponding share of employment, confirming large differentials in productivity between small scale and medium and large firms.

 

 

3.                Conclusions

 

In this paper an attempt has been made to estimate the contribution of informal sector in terms of its employment and GDP with respect to the Pakistani economy. As a direct estimation of the value added of the informal sector is not possible at the moment, an indirect approach is followed using the 'labour input' methodology, which requires data support from both establishments’ and households’ surveys.

 

The LFS is the main source currently used in Pakistan for producing statistics on employment of the informal sector. Like in many other countries, the definition of the informal sector adopted in the Pakistani LFS is not fully harmonized with the Delhi Group recommendations, as the criterion of registration of the enterprise is not implemented and households' employing paid domestic employees are included in the definition. Moreover, the quality of the information on the characteristics of the enterprise (above all the one on accountancy practices) collected from employees is not as accurate as the one coming from self-employed and this is likely to have quite a significant impact on the size of the sector. The main findings of the LFS confirm that large differences in working time, compensation and job turnover exist between formal and informal employment and that wage differentials by sex are higher in the informal sector; at the same time, the results point out that the youngest workers and the less educated are over-represented in this segment of the Pakistan economy, while, contrary to expectation, women are not (but this may be due to underestimation of their participation in the labour market).

 

The major problems concerns the estimates of productivity by industry of the workforce engaged in the informal sector: except for manufacturing, virtually no estimates of value added per worker are currently published, even if proxies could be calculated for some industries on the basis of the existing sources, provided that data broken down by firm size would be made available by the FBS. An attempt has been made to use the corresponding estimates of value added per worker of the Indian economy, appropriately adjusted to take into account that wages in Pakistan are considerably higher than in India. The results of this exercise show that the contribution of the informal sector to total Value added is substantial in Wholesale and Retail Trade and Hotels and Restaurants, accounting for over half of the total value added of the industry, and quite large also in Transport & Communications and in Social, Community & Personal Services. Further work is needed to derive appropriate estimate of productivity from Pakistani data sources and to analyze the new methodology used for the compilation of the revised NA, assessing the extent to which informal activities has already been covered by the revised GDP estimates.

 

 

References

 

Charmes J. (1999), Estimation and Survey methods for the Informal Sector, Proceedings of the Conference of the International Statistical Institute, Istanbul.

 

Charmes J. (2000), The Contribution of Informal Sector to GDP in Developing Countries: Assessment, Estimates, Methods, Orientations for the Future, Fourth Meeting of the Expert (Delhi) Group on Informal Sector Statistics, Geneva, August 2000, International Labour Office, Geneva.

 

Heidi Arboleda and Loh, Meng Kow (2000) Estimation of Contribution of Informal Sector to Gdp on Regular Basis, Fourth Meeting of the Expert (Delhi) Group on Informal Sector Statistics, Geneva, August 2000, International Labour Office, Geneva.

 

Hussmanns, R. (2000), Informal sector surveys: Advantages and limitations of different survey methods and survey designs for the data collection, Fourth Meeting of the Expert Group on Informal Sector Statistics (Delhi Group), Geneva, August 2000.

 

Hussmanns, R. (2002), A labour force survey module on informal employment (including employment in the informal sector) as a tool for enhancing the International comparability of data, Sixth Meeting of the Expert Group on Informal Sector Statistics (Delhi Group), Rio de Janeiro, September 2002.

 

Organisation for Economic Co-operation and Development (2002), Handbook for Measurement of the Non-Observed Economy, OECD, Paris.

 

ILO (1993 a). Statistics of employment in the informal sector. Report for the XVth International Conference of Labour Statisticians, 19-28 January 1993, Geneva.

 

ILO (1993 b). Report of the Conference. Report of the XVth International Conference of Labour Statisticians, 19-28 January 1993, Geneva.

 

Verma, V. (1999), Sample design considerations for informal sector surveys, Proceedings of the Conference of the International Statistical Institute, Istanbul.

 

United Nations (1998), Handbook of National Accounting. Household Accounting: Experiences in the Use of Concepts and Their Compilation. Vol 1: Household Sector Accounts, UN, New York.