At its simplest, the argument is that rapid employment growth has occurred in an environment of lax labor regulation; stiffer regulation will raise labor costs, which is an incentive to employ fewer workers; hence enforcing labour standards will seriously dampen employment growth. Furthermore, intrusive and extensive legislation mostly benefits officials that extort by harassment. What is needed then is "benign neglect" in regulation of small enterprises, even while giving special access to finance and other services.[2]

Based on a labor market framework, the argument is typically elaborated as follows.[3] Even though many of today's poor are employed, they are underemployed; others remain poor because of unemployment. Hence growth in employment is critical to accommodate high unemployment and large underemployment, as well as a growing labor force. This is the best form of unemployment insurance possible; furthermore, reduced unemployment can raise wages because of a higher reservation wage due to a higher expected wage.[4]

Continued low levels of domestic savings and investment in the Pakistan economy, plus a drastic contraction of the public sector, imply that employment growth has to come mainly from the additions and expansion of micro, small and medium enterprises, which are also more labor-intensive than larger enterprises. But the growth of such enterprises may be significantly reduced by the higher labor costs imposed upon business by enforcing (or even requiring) a stiffer labor regulatory framework for micro and small enterprises (MSEs). Private investment may be reduced altogether, or shifted to more profitable ventures which diminish employment creation within Pakistan.[5]

Hence the prospects of rapid employment growth should not be traded off against the expansion of decent work.[6] The need, therefore, is to facilitate growth by diluting existing labor legislation in scope, application, and enforcement. Distortionary policies of government include even minimum wages. Better working conditions, including health and safety, employment security and unemployment benefits, will be delivered by the market when enterprises need them and can afford to do so. A broader social protection system should be used to achieve the objectives of decent work. Hence government should focus on expanding employment, and workers should negotiate terms of employment with enterprises.

The paper reviews this (generic position) in the following manner. It begins with reaffirming the obvious significance of MSEs for labor and economy. This is followed by examining the proposition of a growth-trap for MSEs: higher labor costs would reduce employment and productivity growth by constraining the scale of new enterprises and expansion of existing enterprises. The discussion is confined to non-agriculture, since agriculture involves broader issues of agrarian reforms that lie beyond the scope of this study.[7]

MSE Employment and Output

The main source of published data on employment is the Federal Bureau of Statistics, which conducts the annual Labour Force Surveys. For output data, the main source is the Ministry of Finance, which publishes data on national accounts through the annual Pakistan Economic Survey.[8] An additional source is the Pakistan Statistical Yearbook published by the Federal Bureau of Statistics.[9]

Table 1 provides the most recent published economy-wide data on employment. An approximation to MSE employment is informal sector employment.[10] MSEs account for just under one-half of total country employment (42 million); excluding agriculture, for nearly three-fourths of the remaining employment (24 million). Almost half of all employees hold jobs in MSEs.

Table 1. Structure of Employment in Pakistan, 2003-2004

ƒâ€š Â

Source: Labour Force Survey. It defines informal for non-agricultural employment as including all self-employed; unpaid family labour; employers and employees in household enterprises with less than 10 workers; hence informal employment in larger enterprises is excluded. Younger children are wholly excluded from the survey, regardless of their participation in work. Rural and urban refer to residence of household rather than location of enterprise.

A compelling reference is also to (open) unemployment "" which would have swelled to half the labour force if MSEs had not provided (some kind of) work for these 17 million, mostly men but also women, children and youth, as employees, self-employed and unpaid family labor (and the latter include quasi-employees as dependent upon regular outsourced work by other small enterprises).

Disaggregating employment by economic sectors confirms the pervasiveness of MSEs. Trade, Restaurants & Hotels account for nearly a third of total informal employment. Manufacturing, and Services & Utilities provide an additional third of MSE jobs.

As has often been remarked, the Construction sector is an important source of employment by MSEs; and consequently ranks almost the same as Trade etc., with over 90% as the share of such employment in total employment within the sector. Though Transport & Storage provides an equally smaller (absolute) number of jobs, the nature of the sector is such that it lends itself to informal employment, and therefore the sector ranks third in MSE share of employment.

Deserving of special notice is female employment. The absolute number remains very small in comparison to male employment, reflecting not only much lower labor force participation but also larger unemployment (and underemployment) rate among females. Such data illustrates large deficits in rights to work, serious even with reference to South Asia, or to Islamic Iran and Indonesia.

Contributions of MSEs to non-agricultural production cannot be disaggregated in as much detail as employment. A crude approximation is through sectoral composition of national income accounts. Since (measured) productivity of workers is likely to be lower in MSEs as compared to the average, national income data can overstate the direct contribution of MSEs (though not necessarily that of workers). Since virtually all enterprises consume products and services from MSEs, the indirect contribution of MSEs may, however, be substantial.

To begin with, current contributions can be gauged by focusing on agriculture and the small-scale and household manufacturing sector. Currently (in 2004-2005), these account for over a fourth of GDP "" obviously much less than shares in employment since the arithmetic of market prices does not necessarily capture an alternate social accounting of contributions.[11] Nevertheless, this share reasserts a substantial contribution of MSEs to the economy.

Additional contributions of MSEs are reflected in the GDP generated in other sectors where informality of employment is widespread. Trade now accounts for nearly 20 percent of GDP. Construction and Transport together produce another 20 percent; and Services account for a similar share of GDP. If we recognize the fact that even large-scale manufacturing relies upon inputs from MSEs as well as production through informal employment, then its sectoral contribution is illuminating at over 10 percent of GDP.

National accounts of government allow us to examine one type of objection to a minimum wage in Pakistan. There is an assertion that, unlike in developed countries, minimum wage does or will exceed average income, and therefore privileges some over most. Official data do not appear to support this assertion.

Annual income is estimated by government at Rs 44,000 per capita. After the revision in mid-2005, the current monthly minimum wage (though not received by most workers) is Rs 3,000, which translates to Rs 36,000 annually. Even for a single-person household the wage is less than national average income. Most workers support dependents, and hence current minimum wage per capita becomes substantially less than average income per capita.

Output Growth

Sectoral GDP trends are discussed next. Imputing significance to MSEs requires an examination of growth in those sectors where MSEs provide much employment.[12] Recent growth rates in agriculture, and in small-scale and household manufacturing have exceeded five percent per year. In industry too, growth is officially estimated to have been very good; other sectors are said to have done even better. [13]

Across the three decades, two sectors stand out as better than average growth "" reflected in expanded shares of GDP are Small Scale Manufacturing where the sectoral share has gone up by 50 percent; Transport & Storage, where the share has doubled to become 14 percent of GDP. These are sectors with much employment through MSEs.

An equity perspective is for such rapid growth in production to be recognized as contributions of both labor and capital. Hence it shows much room for redistribution just of the economic growth itself. Minimum labor standards are one way of achieving this redistribution.[14]

Table 2. Sector Shares in Pakistan Gross Domestic Product (%)

Source: Pakistan Economic Survey. GDP is at current factor cost.

Employment Growth

Data suggest that employment expansion has been rapid in MSEs "" official surveys show a steady increase in employment, both absolute and relative, in the informal sector over the last three decades. This has happened despite (at least some) labor legislation, but which has been complied with and enforced more as an exception rather than the rule.

MSEs are pervasive in non-agricultural sectors. These sectors have accounted for more than two-thirds of all additional employment in the economy. More than half of the new jobs came from Trade & Restaurants; Services; and Manufacturing.

Trend data on the informal sector labor force is available since the mid-90s. It indicates that the role of MSEs has grown in non-agricultural employment - from a share of 65% in 1996-97 to 70% in 2003-2004. MSE employment now stands at nearly 40% of all employment.

Since this data excludes informality of employment within the formal sector, the issue of exclusion in labor protection is even more urgent.[15] An idea of this aspect of informalization of employment may be gauged from the fact of massive decline in employment under the Factory Act - across the '90s, such employment nearly halved even as the number of registered and working factories increased, and employment in manufacturing rose by more than 30%.

Surveys of household (exempt from laws) and small manufacturing enterprises (where some laws apply) provide additional insights.[16] Over the decade of 1987-88 to 1996-97, there was a steady growth in the number of units - from about 260,000 to nearly 430,000 enterprises across (rural and urban) Pakistan. Employment expanded at a higher rate, from just above 625,000 to over 1.2 million persons; and paid employees doubled from around 210,000 to nearly 490,000 workers. There is no reason to doubt that, when published, the most recent census of 2002 will establish continuation in these trends.[17]

When the focus is on expanding employment, a central concern is whether labor legislation imposes, or could impose, labor costs that would be a serious constraint upon the growth of employment in MSEs. This can be further decomposed into constraints for setting up new enterprises; and for expanding enterprises. In either case, economies of scale could suggest larger enterprises, which may also have higher productivity that may translate into higher labor compensation for most if not all workers. This expansion may, however, be deterred by the costs of meeting those higher standards of labor protection that are tied to scale of enterprise.[18]

The household and small manufacturing surveys point to scalar growth of such enterprises. Average employment increased (from 2.8 to 3.5 persons) in urban small manufacturing units. Units with eight or more workers more than doubled, and their employment increased nearly four-fold. In contrast, urban units with three or less workers declined in absolute numbers (though slightly).[19]

Table 3. Employment Trends in Pakistan

ƒâ€š Â

Source: Labour Force Survey.

Compliance in MSEs

If MSEs seriously worry about labor costs in considering initiation or expansion, then there should be some evidence of compliance with labor laws that do apply to all enterprises. One example is minimum wages. Recent data from labor force surveys show that over 40% of all employees (across both informal and formal sectors) received less income than the minimum monthly wage (of Rs 2500). Female employees fared worse, with nearly 75% under the minimum wage (in part a reflection of more part-time labor). Urban females fared much worse, but urban males did better. Case studies suggest that informal sector workers would have shown up as doing worse if disaggregated labor force surveys were published.[20]

Data on income is not tied to working hours; hence income larger than minimum wage conceals the violation of minimum wage for maximum working hours and weekly paid leave. Independently of minimum wage, laws on hours and leaves apply to all enterprises (such as the West Pakistan Shops & Establishments Ordinance 1969). However, both across Pakistan and in urban areas, almost 40% of employees worked more than 48 hours in a week. It is doubtful that most received extra payments let alone being paid at legal overtime rates. The situation has been exacerbated by legal extensions in working hours.

A third example is child labor. Official estimates, considered by many to err on the low side (e.g. by underestimating domestic labor, and home-based labour), show growing numbers of children in employment - now of 2.5 million children under the age of 14. MSEs, specially at the lower end of the scale, have been noted in more than one study (including those by the ILO), to employ child labor ignoring laws on special working conditions. Hazardous labor by children is also reported in various studies. Most observers would agree that child labor is an effort to reduce labor costs, in violation of minimum wage law, and of laws against discrimination. Brick kilns are an illustration of larger enterprises where child labor is routine.

Brick kilns also provide a fourth example, that of debt bondage.[21] Laws prohibit forced labor across the board. Yet kilns have grown in number and size, where debt bondage is the rule. Brick kilns are also a specific example of non-compliance with minimum wages leading to bondage. There is evidence that debt bondage is more pervasive than just in kilns, and that it is growing in non-agriculture (even as it declines in agriculture because of the decline in share cropping and field labor).[22]

A fifth example is gender discrimination, supposedly against the law for all enterprises. A specific example is again brick kilns, where piece rates fail to account for the regular labor of women (and of children) when minimum wage is the reference.[23] Aggregate data on wages and on occupations is also an indicator. In 2003-2004, mean income of even urban female employees was around 70% that of males. The comparative distribution of employees further highlights inequality. Even in urban Pakistan, one-third of female employees earned less than Rs 1500 a month, as compared to around one-tenth of males. More than 40% of male employees earned in excess of Rs 4000, while only 25% of females could do as well.

The distribution of female employment across sectors is one reason, suggesting entry barriers for females: among the employed in non-agriculture (which includes self), nearly 40% of females were in the informal sector, as compared to a third of males. Among urban female workers, more than a fourth held unskilled jobs, as compared to less than 20% of males.

In addition to differential access to work, compensation too was highly discriminatory across sectors and occupations. A recent income survey finds urban females employed in "elementary occupations" to earn, on average, less than one-half of males; females as craft workers and service workers fare worse.[24] Among sectors, the difference in income is more than three-fold to the advantage of males for persons employed in manufacturing; and over two-fold for employment in trade.

These examples can be used to infer the low probability of facing higher labor costs due to labor standards, at least at current levels of compliance and enforcement. Of course, one may also infer that MSEs may yet hesitate at expansion because larger scale increases "visibility," and hence a higher probability of application of higher labor standards, with higher costs of compliance or of evasion.

It is simplistic to conclude that these must remain necessary social costs of rapid growth in employment; and reasonable to insist that MSEs should and can do better in the future.

Some light may also be thrown by the surveys of manufacturing on whether labor costs have impeded scalar growth of MSEs. Among urban small manufacturing, enterprises with three workers or less were stagnant in the numbers employed, over the period 1987-88 to 1996-97. For this scale of enterprises, labour costs were less than 10% of production value in the recent year. In contrast, employment in relatively larger enterprises with eight or more workers grew almost five-fold (some of which may be conversion from household units to small enterprises), where labor costs were twice as high at around 20% of production costs. At least for these scales of enterprises, labor costs did not deter growth. Of course, one inference could be that if growth is to be maintained labor costs should not increase. But labor costs as a share of production costs were much lower (at 15%) in the earlier year, at least for the larger units with eight or more workers.

Figure 1. Growth in Urban Small Manufacturing in Pakistan

Source: Survey of Small and Household Manufacturing Industries. Employment includes owners, unpaid family labour and paid employees. In 1996-97, the survey includes 10 or more persons in unregistered enterprises.

Compliance in Larger Enterprises

A counter-factual could be asserted - that employment growth in MSEs would have been even more explosive than it actually was, because of costs imposed by labor legislation (as deterrence, or as reduction in investible surplus). So relatively larger enterprises were not established as often, or they were less labor-intensive than they should have been. It seems to us that a complex set of issues are involved.

One way of examining this proposition is to look at compliance data of larger enterprises. Smaller enterprises may be expected to seriously worry about legislated labor costs if they observe a high level of compliance among relatively larger enterprises; infer that similar labor costs will in fact apply to them, and these labor costs are considered to be "excessive."

Most labor legislation has applied to enterprises with 10 or more (paid) workers.[25] This is usually interpreted by enterprises, and complicit labor officials, to mean full-time employees with formal written appointments. Hence the first thing to note is that micro enterprises may grow substantially in scale, and a vast range of scales may be available for a new micro enterprise, without worrying about most labor legislation.

Using the latest data available, for 1996-97 the average employment (including the working owner and unpaid family members) in household and small manufacturing units was less than four persons; around 90% had four or less workers.[26] Less than half of all workers were paid employees. Hence most micro enterprises in manufacturing could have more than doubled employment without concern for the costs of most legislated labor benefits, and even of the infamous Factory Act per se. The latter has little teeth, anyway "" in 2001, less than 10,000 units were registered across all of Pakistan; of these less than 2,000 were considered as "reporting factories."

The option of expanding wage employment within the enterprise has been independent of the possibility of employing workers in other arrangements such as piece rate and contract, which is usually excluded in the minimum number of workers for legislative obligations. For example, in 2001-2002 the average daily number of factory workers paid under the Payment of Wages Act was around 160,000 persons, as compared to well over two million persons reported to be employed outside of the informal manufacturing sector.

Compliance data for larger enterprises does not support the contention that small enterprises be concerned with labor costs when crossing the boundary of 10 workers. In 1998-99, around 700,000 non-agricultural workers were 'secured' by provincial social security institutions; coverage remained well under a million workers by 2002-2003.[27] In the latter year, formal sector employment in non-agriculture was around well over seven million persons; even conservatively restricting the reference to the urban formal sector (4 million), most of the employees (over 2.5 million) were excluded. Extension of social security obligations to all enterprises with five or more workers has shown no corresponding expansion in the number of contributing enterprises. One reason for this will of course be the virtual disappearance of inspections as policy.

Coverage for other labor welfare levies remains inadequate. The Employees Old-Age Benefits Institution had registered only around 55,000 employers by 2002-2003, covering less than two million employees.[28] Consider the Workers Welfare Fund, applicable to industrial establishments. One cess is (at 2% of) income of all establishments with "assessed" income of Rs 100,000 or more. Another applies to the profits of those with 50 or more workers, clearly irrelevant to small enterprises. Data is unavailable for enterprises contributing and workers covered.

Other aspects of labor welfare may well worry MSEs. Formation of trade unions and consequent collective bargaining for better labor protection is one. But data on union membership suggest otherwise, with less than 1,000 reporting unions and less than 300,000 members in recent years.

Termination costs in particular, and generally checks upon arbitrary hire and fire, could well constrain MSEs from expanding to a size where such laws constrain "labor flexibility." These costs may be thought unaffordable for small enterprises. It isn't quite obvious where the level lies, but balance is to be struck in the sharing of unemployment and retraining costs, risk, uncertainty and income, between workers, enterprises and the state.[29]

Figure 2. Output and Employment Growth in Pakistan Non-Agricultural Sectors

ƒâ€š Â

Source: Labour Force Survey; Pakistan Economic Survey. Output is GDP at constant prices of 1980-81Another part may be subsidised by government in the same manner as it does, or will, subsidise capital and other input costs for MSEs.38 Even owners of expanding MSEs may well have to trim their profits if their product markets do not expand sufficiently. What government can then do is to help with the adjustment, which includes reducing overheads of compliance that include dealing with enforcement organisations. Furthermore, labour welfare costs should and can be more equitably shared by workers, and among enterprises. Some suggestions in this regard are given in later sections. What government should discontinue doing is to shift costs entirely upon some workers by exemptions, e.g. in special industrial zones.

Field Assessments

Discussions were held with selected workers, enterprise owners, their associations, and government officials in Karachi, Hyderabad, Faisalabad, Lahore and Islamabad. Since a systematic survey of enterprises was not called for by the study, selections were on the basis of credible contacts. By design, agriculture and fisheries were excluded. Both home-based workers and domestic workers also require separate studies by themselves.

The selection ended up being more at the lower end of MSEs. Enterprises and workers included the following: garments; bangles; power-looms; brick kilns; light engineering; automotive parts; rickshaw, taxi and bus; cartons; tea; and school. Interviews with government included those with the federal Workers Welfare Fund; Punjab Social Security and Punjab Minimum Wage Board; and District Labour Officer.[30] SMEDA was not included in the interviews because much written materials were available.

These discussions were supplemented by discussions at, and proceedings of, three recent country conventions of (mostly) informal economy worker associations in Karachi (Textiles and Transport, which included self-employed) and Multan (Construction, mostly brick kilns).

Hence, assessments should be treated as impressions rather than "hard" data. On the basis of extensive work done previously by PILER, these impressions are more robust than indicated by the special sample itself.[31]

Assessments include these findings:

• Enterprise owners do not consider labor costs as a significant constraint in setting up or growth of enterprises; other input costs are a source of worry[32];
• One reason for this is that most MSEs do not comply with most labor laws; including minimum wage and social security provisions; many are not even aware of legal requirements; inspection is seen only as benefiting inspectors rather than workers;
• Workers are keen on adequate wages and regular payment; this is a major reason for wanting written terms of appointment generally denied by owners in MSEs;
• Social security is keenly desired by workers, but if conveniently accessible and independent of employer certification for enrolment and benefits; health care is specially desired by workers susceptible to serious injury at work;
• Employers are averse to collective bargaining whereas workers are keen on broad sectoral unions which could balance employer power at the establishment level[33];
• Government does little to enforce laws in general; and labor laws, even minimum wages, are no exception;
• The recent change from inspection to self-assessment has yet to show positive results; and
• A federal institution for oversight of labor welfare would improve effectiveness - both as more inclusive, as well as more cost-effective.

Required Reforms

With present legislation, micro enterprises can expand quite a bit, exempted as they are from formal compliance requirements. For small enterprises that expand or establish to enter the formal sector, enforcement is so lax that they need worry little about compliance costs; avoidance through outsourcing and contracting is a feasible cheaper alternative. Both will allow them to maintain or increase profits.

Hence, often asserted is the reverse. Instead of legislation being a brake on MSEs, informalization of employment, including MSEs, has grown more rapidly than otherwise so as to avoid and evade labor legislation applicable to the formal sector. There is no disputing this global phenomenon. One response could be to dump most labor legislation, so that enterprises can expand or larger enterprises can be established without worrying about additional obligations of labor legislation.

For those committed to a society with decent work, the more appropriate response is to acknowledge that MSEs are where most workers are now located; hence, to modify legislation towards (facilitation and enforcement) of universal application of minimum labor standards. There are just too many observed market failures to establish and maintain minimum standards to suggest that the state can withdraw from effective regulation.

Often expressed is the view that MSEs offer an escape valve (directly or indirectly) for enterprises wishing to avoid compliance costs of labor legislation. When standards are universally applied and enforced, labor costs should become less of an issue in choosing an appropriate scale for competitiveness.

What these universal standards should be are an issue for tripartite agreements (or social dialogue) "" except for the core labor rights in the Declaration. Even in the non-agricultural sector, the "labor market" falls short on one or more of these standards - child labor, debt bondage (forced labor and more), discrimination (gender and beyond), collective bargaining (enterprise and more). Furthermore, realizing these rights entails recognition of complementary standards, such as adequate wages for all adults, men and women. The concluding section deals with these issues in more detail.

Naturally this will entail costs to MSEs that owners (and customers) have been shifting so far upon workers and citizens. What these costs will be is a matter of prior, tripartite consensus upon specific minima as universal labor standards, and upon cost-effective compliance and enforcement mechanisms. This must include adequate inspection arrangements as long as worker organisations remain weak in comparison to even individual employers. But the latest IRO 2002 is an example of additional constraints imposed upon collective organization and action by workers, and hence the increasing obligation upon the state to seek compliance with laws for labor welfare.

Present legislation, including the drafts being considered as consolidated legislation, remain irrelevant to a costing exercise since this legislation is not aimed at universal application.[34] As they stand, most will not, or cannot, be enforced universally. The experience with the law against bonded labor is also not encouraging - where the government took several years to even make rules for implementation. There are other issues, dealt with in subsequent sections.

Are there any labor standards that impose virtually no cost upon enterprises, at least not as directly as a minimum wage, or social security levy, or the replacement of child workers with adults? Clearly, recognition of trade unions and collective bargaining is one. Of course, successful bargaining will lead to higher future costs of meeting other labor standards.

Some idea of prevailing labor costs may help in reducing resistance to reforms. Delay in publication of the Economic Census necessitates the use of much older and narrower data, for small manufacturing units in 1996-97.[35]

As the scope for bearing higher labor costs, consider the average share of labor costs in urban production and in total costs, which was 15%, and 25%, respectively.[36] At the upper end of the employment scale, in enterprises with 8-9 workers, these shares rose to around 20% and 30%; declining to about 25% of production in enterprises with 10 or more workers. The large, perhaps considerable, profit margin (inclusive of compensation for working owners, and taxes) is indicated by the relative gap between production and costs.[37] On average, this was nearly 50%; declining to around 30% among the largest units.

Another way of looking at the scope for increasing compensation is productivity of workers. The large gap between growth of national output and of employment is an indicator, though crude. Unfortunately, trend data on real wages remains scanty. When official estimates of unskilled construction wage are adjusted by the consumer price index, the real wage in Lahore fell between 1993-94 and 2002-2003 - a period over which national output in construction rose by 15 per cent.

Whatever the reforms to encompass MSEs within labor legislation, labor costs are likely to rise, and profit levels may even decline though less so when scalar expansion is also involved. Some enterprises may even shut down, as they should if survival depends upon exploitation, or even upon forcing labor to work for inadequate compensation just because owners choose to be (independent) employers rather than (disciplined) wage workers.

When profits are seriously affected, policy makers need to decide whether profit rates or profit levels are the object of concern. One approach to persuasion could be the argument that if the scale of profits increases with scale of enterprise, it is reasonable of workers and society to hold the larger enterprise accountable for larger contributions to labor standards, even while workers remain grateful for the additional employment.

With reductions in underpaid, overworked, and insecure workers, an enterprise could well gain through higher labor productivity a large part of what it pays in higher labor costs (as surveys show, and more than one ILO study argues). Some parts of additional labor costs will get shifted through higher prices upon consumers (including those now subsidized in affluent importing countries).

Some would also suggest that the burden of labor protection be shifted away from enterprises to mainly upon government through comprehensive measures for social protection. Indeed, some aspects could be handled more cost-effectively, at least in principle, by a more capable government. However, some are inescapably the responsibility of the enterprise, e.g. minimum wages without discrimination, safe and healthy work environment including working hours and collective bargaining. As for social protection, collective efforts of workers and enterprises, and other alternatives, could prove more cost-effective than excessive reliance upon government.

In the event that expansion of employment does seem to still be strongly threatened by higher labour costs for MSEs, then better macro management of the economy is required "" which would follow from a vision of development much unlike that currently followed by the state (or advised by donors).[39] Hence it is important to stress that government can do much more when the state expresses a greater commitment to citizen rights.[40]

To conclude, it is our judgement that neither official survey data nor field assessments support the fear that realizing minimum labor standards for decent work will become a serious constraint to employment growth in MSEs. Social norms cause deficits in decent work; changing these norms would be the core objective of a democratic government.

* This paper is based on a study prepared for the ILO, as part of the Research, Education & Training program supported at PILER by EU, NOVIB, FNV, Stichting De Zaaire and Frere des Hommes. None of these organizations are responsible for the positions taken in the paper. Comments are encouraged at piler@cyber.net.pk.

Endnotes:

1 The Country Assistance Strategy documents of the Banks provide an introduction to the advice given to Pakistan. The approach is given in the Technical Assistance papers that precede loan documents, and detailed advice is provided in the loan documents.

2 E.g. the UNDP in Small and Medium Enterprise Development. The arguments for relaxing environmental standards for MSEs are often the same "" as the "price" of growth, even if the price is mostly borne by the poor rather than those with a greater capacity to bear it.

3 An illustration is the World Development Report: A Better Investment Climate for Everyone. Additional references are provided at the end of the paper.

4 Expected wage equals probability of employment times the prevailing wage.

5 Lowering the attractiveness of these alternative ventures is not apparently considered, at least seriously.

6 The ADB has formally endorsed the Declaration but the WB has not. Effectively it has made little difference in operations of the two Banks within Pakistan.

7 There can be no disagreement that effective enforcement of essential labor standards in the non-agricultural economy depends much upon agrarian reforms. Minimum wages is the obvious example; so is debt bondage. One can only take the familiar plea of constraints given by the study TOR and corresponding resources. Some issues have been discussed in PILER studies on debt bondage, and in its comments on an ADB loan for reducing rural poverty in Sindh.

8 The Social Policy & Development Centre (SPDC) maintains critical views of output growth data. The State Bank attempts "independent" estimates; but its autonomy remains constrained structurally, domestically, and internationally by affiliation of leadership with the IMF, WB, and now likely with the IFC or ADB.

9 All data used in the paper are from one or more of these sources unless an alternative citation is given.

10 "Informal sector" is defined by the Federal Bureau of Statistics by nature and scale of employment (Table 1). It is not completely informal since some labor (and tax) laws apply to all enterprises, but compliance is the exception. Of course, distinction between informal and formal is becoming ambiguous with the formal sector itself increasing numbers of such employment as is excluded de jure or de facto from application of labor legislation.

11 As an easily understood example, the puzzled reader may think of differences between low domestic and high international prices for commodities and services, including labor.

12 This is particularly important in implementing labor standards, since enterprise owners in stagnant sectors are unlikely to be keen on rights of workers other than themselves.

13 Some scepticism is even expressed by the ADB in the recent country Economic Update.

14 Even among neoliberal economists, only the very naive would assert that production and distribution can be managed independently as models in introductory textbooks in mainstream economics. For a powerful critique, see Employment, social justice and societal well-being 2002.

15 ILO's Economic Security for a Better World, suggests the following "status" dimensions for defining empirically a continuum of informal labor: regularity; contract; workplace; employment protection; social protection.

16 Survey of Small and Household Manufacturing Industries 1988 and Survey 1996-97.

17 Publication has been delayed, but a request for data was refused by the Federal Bureau of Statistics.

18 As incentive to expansion, standards may then be reduced. Or they may be made universal, doing away with reverse incentives to remain small. There may also be other options for (explicitly) sharing costs of labor protection with the enterprise, in order to maintain minimum labor standards.

19 Note also that employment in small units (487,000) remained many times larger than in household units (85,000); despite de facto; if not de jure, exemptions of household units from labour legislation.

20 Subcontracted Women Workers in the World Economy: The Case of Pakistan; Women's Work & Empowerment Issues in an Era of Economic Liberalisation: A Case Study of Pakistan's Urban Manufacturing Sector.

21 Unfree labour in South Asia: Debt bondage in brick kilns; Bonded Labour in Pakistan: An Overview.

22 Rapid Assessment Studies of Bonded Labour in Different Sectors in Pakistan.

23 Op.cit.

24 Household Integrated Economic Survey 2001-2002. Unfortunately, published data provides information only for all employed, which includes self-employed and unpaid family labor.

25 Labour Market Policies and Institutions: A Framework for Social Dialogue. See the discussion on draft legislation for a (partial) list of laws currently applicable to medium and large enterprises.

26 Survey of Small and Household Manufacturing Industries 1996-97.

27 Report of the Task Force on Labour Welfare Levies; Combating Poverty: Is Growth Sufficient? Beyond Safety Nets: Provision of Comprehensive Social Security in Pakistan.

28 Combating Poverty, op.cit.

29 See Economic Security for a Better World; Employment, social justice and societal well-being.

30 Interviews at the federal Ministry of Labour with the Central Labour Advisor and at the federal Planning Commission with the Secretary, Planning Commission or its Chief Economist could not take place due to unavailability of respondents; so was the case for the Chairman, Punjab Planning & Development.

31 And are certainly not anecdotes as feared by a reviewer.

32 Similar findings are reported in Encouraging New Growth Areas and Employment, a study for the ILO at the Mohammed Ali Jinnah University, Islamabad; and in SME Development in Pakistan: Analysing the Constraints to Growth, a study at the Lahore University of Management Sciences (LUMS) for the ADB. As would be expected, "business leaders" gave more importance to "labor problems" than the enterprise owners themselves. This is not unlike elsewhere in South Asia; see Labour processes & productivity in micro and small enterprises.

33 An interesting initiative has been taken by power loom workers in Swat. Their union has inducted workers from other sectors, even wholly unrelated, such as workers in cosmetics manufacturing. Negotiations are now carried out with a broad range of employers directly without any reference to labor officials who are seen as irrelevant or an obstruction to negotiations.

34 A companion paper discusses draft legislation. Only the Industrial Relations Ordinance has been promulgated. After severe criticism by workers, government and opposition parties have made some noises about amendments. The review and consolidation of labour legislation began when Omar Asghar Khan was the federal Minister for Labour; no progress has been apparent after his death.

35 Survey of Small and Household Manufacturing Industries.

36 The Survey defines production value to include indirect taxes, but cost does not include them.

37 A rough idea of owner-worker compensation can be had from the single-person enterprise in the official survey: around Rs 3500 per month, as the difference between production and costs. Average monthly employment cost was around Rs 1000 per worker; at the upper end of the enterprise scale it was almost twice the average.

38 Subsidies to capital in property and in person come in various forms. Quite concealed but enormous are state security expenditures as one example. Many examples of infrastructure can be given, but two should suffice: Airports and Motorways where user charges are a fraction of costs; and the exemptions of property taxes to elite private schools and hospitals. The scope of public outlays for labor welfare is enormous, of which just one example is the enormous military budget.

39 For a general discussion, see Employment, social justice and societal well-being; Circumventing macroeconomic conservatism: A policy framework for growth, employment and poverty reduction. In Pakistan, there are various critiques of the PRSP by civil society, e.g. Combating Poverty: Is Growth Sufficient?

40 For example, government has some room for budget reallocations; but serious reformulation of budgets requires cuts in military expenditures, deletion of tax exemption for agricultural incomes, and reduction in external debt. Last year, just the revision in budgeted defence spending was an additional Rs 22 billion, which is many multiples of the total (enterprise and government) funding for Social Security, EOBI, and Workers Welfare Fund in recent years. Human security versus state security remains a central issue in governance in Pakistan.

References:

Asian Development Bank, Country Strategy & Program Update 2006-2008.

___________Country Strategy & Program Update 2004-2006.

___________Developing Social Health Insurance Project (TAR:PAK 37359).

___________Enhancing Capacity For Resource Management and Poverty Reduction In Punjab (TAR:PAK 36057).

___________Labour Markets in Asia: Promoting Full, Productive, and Decent Employment 2005.

___________Labour Reforms and ADB's Role in Pakistan Small and Medium Size Enterprise 2003, speech by ADB Country Director,

___________ Preparing The Social Protection Strategy (TAR:PAK 37008, TA4155-PAK);

___________ Punjab Resource Management Program Loan (RRP:PAK 36057).

___________Punjab Local Justice Support Program.

___________Punjab Resource Management Program 2 (TAR:PAK 37202).

___________Social Protection Capacity Building in Pakistan.

___________ SME Development in Pakistan: Analysing the Constraints to Growth 2005.

___________SME Sector Development Program Loan (PAK RRP: 34327).

___________SME Sector Development Project Loan (PAK 34327-02).

___________Small & Medium Enterprise Trade Enhancement Finance Project Loan (RRP:PAK 33021).

Economic & Political Weekly, ADB: case study of Access to Justice Programme 2002.

___________Unfree labour in South Asia: Debt bondage in brick kilns 2004.

Government of Pakistan, Constitution of the Islamic Republic of Pakistan 1973.

___________Industrial Relations Ordinance 2002.

""""""Federal Bureau of Statistics, Household Integrated Economic Survey 2001-2002

___________Pakistan Social and Living Standard Measurement Survey 2004.

___________Pakistan Statistical Yearbook 2004.

___________Survey of Small and Household Manufacturing Industries 1988.

___________Survey of Small and Household Manufacturing Industries 1996-97.

______Ministry of Finance, Accelerating Economic Growth and Reducing Poverty: The Road Ahead 2003.

___________Pakistan Economic Survey, various years.

___________Recommendations of the Task Force on Pension Funds 1996.

___________Report of the Task Force on Labour Welfare Levies 2000.

______Ministry of Industries, Towards a Prosperous Pakistan: A Strategy for Rapid Industrial Growth 2005, draft.

______Ministry of Labour, Final Report on Task Force on Labour Welfare 2001.

__________ Labour Protection Policy 2005, draft.

__________ Rapid Assessment Studies of Bonded Labour in Different Sectors in Pakistan 2004.

___________Conditions of Employment Ordinance, draft.

___________Payment of Wages Ordinance, draft.

___________Occupational Safety & Health Ordinance, draft.

___________Human Resources Development Ordinance, draft.

___________Labour Welfare & Social Security Ordinance, draft.

______Planning Division, Medium-Term Development Strategy 2005-2010.

______Small & Medium Enterprise Development Authority [SMEDA], Developing SME Policy in Pakistan: SME Issues Paper n.d.

__________ Recommendations made by the Working Committees of the SME Task Force 2004.

ILO, Creating a Conducive Policy Environment for Micro, Small and Medium-Sized Enterprises in Pakistan 2002.

___________Debt Bondage in Pakistan: Making Law Matter 2002.

___________Economic Security for a Better World 2004.

___________A Fair Globalisation: Creating Opportunities for All 2004.

___________A Future without Child Labour 2002.

___________A Global Alliance against Forced Labour 2005.

___________Organising for Social Justice 2004.

___________Normalised and Disaggregated Gaps in Workers Rights 2003.

___________ Bonded Labour in Pakistan: An Overview 2001.

___________Productivity, Decent Employment and Poverty: Conceptual and Practical Issues Related to Small Enterprises 2005.

___________Small Enterprises, Big Challenges: A Literature Review on the Impact of the Policy Environment 2002.

___________Social Protection: A Life Cycle Continuum Investment For Social Justice, Poverty Reduction and Sustainable Development 2003

___________Stopping Forced Labour 2001.

___________Time for Equality at Work 2003.

___________Unfree Labour in Pakistan: Debt and Bondage in Brick Kilns 2004.

___________Using Small Enterprise Development to Reduce Child Labour 2004.

___________Your Voice at Work 2000.

Indian Journal of Labour Economics, Labour processes & productivity in micro and small enterprises 2004.

Institute of Development Studies, Developing Rights 2005.

___________ Making Law Matter 2001.

International Institute for Labour Studies, Social Protection, Decent Work and Development 2004.

International Labour Review, Circumventing macroeconomic conservatism: A policy framework for growth, employment and poverty reduction 2005.

___________ Discrimination and equality at work: A review of concepts 2003.

___________ Employment, social justice and societal well-being 2002.

___________Equality at work 2003.

___________Globalisation and decent work policy: Reflections upon a new legal approach 2004.

___________Labour law and new forms of corporate organization 2005.

___________Work and Rights 2000.

Mohammed Ali Jinnah University, Encouraging New Growth Areas and Employment 2005,.study for the ILO.

The Network, Industrial Relations and Labour Rights in Pakistan 2004.

Oxford University Press, Labour Legislation & Trade Unions in Pakistan 2001.

Pakistan Development Review, A study on unemployment in Pakistan - Labour Market Dynamics in Pakistan 2002,

PILER, Beyond Safety Nets: Provision of Comprehensive Social Security in Pakistan 2000.

___________Labour Market Policies and Institutions: A Framework for Social Dialogue 2000.

___________Mitigation and Abolition of Bonded Labour: Policy, Law & Economy in Pakistan 2001.

___________Subcontracted Women Workers in the World Economy: The Case of Pakistan 2000.

___________Women's Work & Empowerment Issues in an Era of Economic Liberalisation: A Case Study of Pakistan's Urban Manufacturing Sector 2001.

PILDAT, Understanding Labour Issues in Pakistan 2005.

SPDC, Combating Poverty: Is Growth Sufficient? 2004

UNDP, Small and Medium Enterprise Development 1999.

WIEGO, Learning from Experience: A Gendered Approach to Social Protection for Workers in the Informal Economy 1999.

World Bank, Attacking Poverty 2001.

___________A Better Investment Climate for Everyone 2005.

___________Building Institutions for Markets 2002.

___________Concept Note for the Proposed Labour Market Study 2005.

___________Country Assistance Strategy Progress Report 2004.

___________Generating Investment and Employment Opportunities In Pakistan through Labour Market Reforms.2004.

___________Investment Climate Assessment 2003.

___________Poverty Reduction Support Credit Project: Simplified Implementation Report 2005

___________SMEs, Growth and Poverty 2004.

___________ Workers in an Integrating World 1995.

| More...